id
stringlengths 9
16
| submitter
stringlengths 3
64
⌀ | authors
stringlengths 5
6.63k
| title
stringlengths 7
245
| comments
stringlengths 1
482
⌀ | journal-ref
stringlengths 4
382
⌀ | doi
stringlengths 9
151
⌀ | report-no
stringclasses 984
values | categories
stringlengths 5
108
| license
stringclasses 9
values | abstract
stringlengths 83
3.41k
| versions
listlengths 1
20
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-11 00:00:00
| authors_parsed
sequencelengths 1
427
| prompt
stringlengths 166
3.49k
| label
stringclasses 2
values | prob
float64 0.5
0.98
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1603.01684 | Chenxing Xia | Hanling Zhang and Chenxing Xia | Saliency Detection combining Multi-layer Integration algorithm with
background prior and energy function | 25 pages, 8 figures. arXiv admin note: text overlap with
arXiv:1505.07192 by other authors | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose an improved mechanism for saliency detection.
Firstly,based on a neoteric background prior selecting four corners of an image
as background,we use color and spatial contrast with each superpixel to obtain
a salinecy map(CBP). Inspired by reverse-measurement methods to improve the
accuracy of measurement in Engineering,we employ the Objectness labels as
foreground prior based on part of information of CBP to construct a
map(OFP).Further,an original energy function is applied to optimize both of
them respectively and a single-layer saliency map(SLP)is formed by merging the
above twos.Finally,to deal with the scale problem,we obtain our multi-layer
map(MLP) by presenting an integration algorithm to take advantage of multiple
saliency maps. Quantitative and qualitative experiments on three datasets
demonstrate that our method performs favorably against the state-of-the-art
algorithm.
| [
{
"version": "v1",
"created": "Sat, 5 Mar 2016 06:12:44 GMT"
}
] | 2016-03-16T00:00:00 | [
[
"Zhang",
"Hanling",
""
],
[
"Xia",
"Chenxing",
""
]
] | TITLE: Saliency Detection combining Multi-layer Integration algorithm with
background prior and energy function
ABSTRACT: In this paper, we propose an improved mechanism for saliency detection.
Firstly,based on a neoteric background prior selecting four corners of an image
as background,we use color and spatial contrast with each superpixel to obtain
a salinecy map(CBP). Inspired by reverse-measurement methods to improve the
accuracy of measurement in Engineering,we employ the Objectness labels as
foreground prior based on part of information of CBP to construct a
map(OFP).Further,an original energy function is applied to optimize both of
them respectively and a single-layer saliency map(SLP)is formed by merging the
above twos.Finally,to deal with the scale problem,we obtain our multi-layer
map(MLP) by presenting an integration algorithm to take advantage of multiple
saliency maps. Quantitative and qualitative experiments on three datasets
demonstrate that our method performs favorably against the state-of-the-art
algorithm.
| no_new_dataset | 0.947527 |
1603.04522 | Yong Liu Stephen | Yong Liu, Peilin Zhao, Xin Liu, Min Wu, Xiao-Li Li | Learning Optimal Social Dependency for Recommendation | 8 pages, 2 figures | null | null | null | cs.IR cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Social recommender systems exploit users' social relationships to improve the
recommendation accuracy. Intuitively, a user tends to trust different subsets
of her social friends, regarding with different scenarios. Therefore, the main
challenge of social recommendation is to exploit the optimal social dependency
between users for a specific recommendation task. In this paper, we propose a
novel recommendation method, named probabilistic relational matrix
factorization (PRMF), which aims to learn the optimal social dependency between
users to improve the recommendation accuracy, with or without users' social
relationships. Specifically, in PRMF, the latent features of users are assumed
to follow a matrix variate normal (MVN) distribution. The positive and negative
dependency between users are modeled by the row precision matrix of the MVN
distribution. Moreover, we have also proposed an efficient alternating
algorithm to solve the optimization problem of PRMF. The experimental results
on real datasets demonstrate that the proposed PRMF method outperforms
state-of-the-art social recommendation approaches, in terms of root mean square
error (RMSE) and mean absolute error (MAE).
| [
{
"version": "v1",
"created": "Tue, 15 Mar 2016 01:41:11 GMT"
}
] | 2016-03-16T00:00:00 | [
[
"Liu",
"Yong",
""
],
[
"Zhao",
"Peilin",
""
],
[
"Liu",
"Xin",
""
],
[
"Wu",
"Min",
""
],
[
"Li",
"Xiao-Li",
""
]
] | TITLE: Learning Optimal Social Dependency for Recommendation
ABSTRACT: Social recommender systems exploit users' social relationships to improve the
recommendation accuracy. Intuitively, a user tends to trust different subsets
of her social friends, regarding with different scenarios. Therefore, the main
challenge of social recommendation is to exploit the optimal social dependency
between users for a specific recommendation task. In this paper, we propose a
novel recommendation method, named probabilistic relational matrix
factorization (PRMF), which aims to learn the optimal social dependency between
users to improve the recommendation accuracy, with or without users' social
relationships. Specifically, in PRMF, the latent features of users are assumed
to follow a matrix variate normal (MVN) distribution. The positive and negative
dependency between users are modeled by the row precision matrix of the MVN
distribution. Moreover, we have also proposed an efficient alternating
algorithm to solve the optimization problem of PRMF. The experimental results
on real datasets demonstrate that the proposed PRMF method outperforms
state-of-the-art social recommendation approaches, in terms of root mean square
error (RMSE) and mean absolute error (MAE).
| no_new_dataset | 0.942981 |
1603.04614 | Qingqun Ning | Qingqun Ning, Jianke Zhu, Zhiyuan Zhong, Steven C.H. Hoi, Chun Chen | Scalable Image Retrieval by Sparse Product Quantization | 12 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fast Approximate Nearest Neighbor (ANN) search technique for high-dimensional
feature indexing and retrieval is the crux of large-scale image retrieval. A
recent promising technique is Product Quantization, which attempts to index
high-dimensional image features by decomposing the feature space into a
Cartesian product of low dimensional subspaces and quantizing each of them
separately. Despite the promising results reported, their quantization approach
follows the typical hard assignment of traditional quantization methods, which
may result in large quantization errors and thus inferior search performance.
Unlike the existing approaches, in this paper, we propose a novel approach
called Sparse Product Quantization (SPQ) to encoding the high-dimensional
feature vectors into sparse representation. We optimize the sparse
representations of the feature vectors by minimizing their quantization errors,
making the resulting representation is essentially close to the original data
in practice. Experiments show that the proposed SPQ technique is not only able
to compress data, but also an effective encoding technique. We obtain
state-of-the-art results for ANN search on four public image datasets and the
promising results of content-based image retrieval further validate the
efficacy of our proposed method.
| [
{
"version": "v1",
"created": "Tue, 15 Mar 2016 09:53:32 GMT"
}
] | 2016-03-16T00:00:00 | [
[
"Ning",
"Qingqun",
""
],
[
"Zhu",
"Jianke",
""
],
[
"Zhong",
"Zhiyuan",
""
],
[
"Hoi",
"Steven C. H.",
""
],
[
"Chen",
"Chun",
""
]
] | TITLE: Scalable Image Retrieval by Sparse Product Quantization
ABSTRACT: Fast Approximate Nearest Neighbor (ANN) search technique for high-dimensional
feature indexing and retrieval is the crux of large-scale image retrieval. A
recent promising technique is Product Quantization, which attempts to index
high-dimensional image features by decomposing the feature space into a
Cartesian product of low dimensional subspaces and quantizing each of them
separately. Despite the promising results reported, their quantization approach
follows the typical hard assignment of traditional quantization methods, which
may result in large quantization errors and thus inferior search performance.
Unlike the existing approaches, in this paper, we propose a novel approach
called Sparse Product Quantization (SPQ) to encoding the high-dimensional
feature vectors into sparse representation. We optimize the sparse
representations of the feature vectors by minimizing their quantization errors,
making the resulting representation is essentially close to the original data
in practice. Experiments show that the proposed SPQ technique is not only able
to compress data, but also an effective encoding technique. We obtain
state-of-the-art results for ANN search on four public image datasets and the
promising results of content-based image retrieval further validate the
efficacy of our proposed method.
| no_new_dataset | 0.947186 |
1507.03325 | Zhao Zhang | Zhao Zhang, Kyle Barbary, Frank Austin Nothaft, Evan Sparks, Oliver
Zahn, Michael J. Franklin, David A. Patterson, Saul Perlmutter | Scientific Computing Meets Big Data Technology: An Astronomy Use Case | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scientific analyses commonly compose multiple single-process programs into a
dataflow. An end-to-end dataflow of single-process programs is known as a
many-task application. Typically, tools from the HPC software stack are used to
parallelize these analyses. In this work, we investigate an alternate approach
that uses Apache Spark -- a modern big data platform -- to parallelize
many-task applications. We present Kira, a flexible and distributed astronomy
image processing toolkit using Apache Spark. We then use the Kira toolkit to
implement a Source Extractor application for astronomy images, called Kira SE.
With Kira SE as the use case, we study the programming flexibility, dataflow
richness, scheduling capacity and performance of Apache Spark running on the
EC2 cloud. By exploiting data locality, Kira SE achieves a 2.5x speedup over an
equivalent C program when analyzing a 1TB dataset using 512 cores on the Amazon
EC2 cloud. Furthermore, we show that by leveraging software originally designed
for big data infrastructure, Kira SE achieves competitive performance to the C
implementation running on the NERSC Edison supercomputer. Our experience with
Kira indicates that emerging Big Data platforms such as Apache Spark are a
performant alternative for many-task scientific applications.
| [
{
"version": "v1",
"created": "Mon, 13 Jul 2015 04:47:04 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Mar 2016 19:04:09 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Zhang",
"Zhao",
""
],
[
"Barbary",
"Kyle",
""
],
[
"Nothaft",
"Frank Austin",
""
],
[
"Sparks",
"Evan",
""
],
[
"Zahn",
"Oliver",
""
],
[
"Franklin",
"Michael J.",
""
],
[
"Patterson",
"David A.",
""
],
[
"Perlmutter",
"Saul",
""
]
] | TITLE: Scientific Computing Meets Big Data Technology: An Astronomy Use Case
ABSTRACT: Scientific analyses commonly compose multiple single-process programs into a
dataflow. An end-to-end dataflow of single-process programs is known as a
many-task application. Typically, tools from the HPC software stack are used to
parallelize these analyses. In this work, we investigate an alternate approach
that uses Apache Spark -- a modern big data platform -- to parallelize
many-task applications. We present Kira, a flexible and distributed astronomy
image processing toolkit using Apache Spark. We then use the Kira toolkit to
implement a Source Extractor application for astronomy images, called Kira SE.
With Kira SE as the use case, we study the programming flexibility, dataflow
richness, scheduling capacity and performance of Apache Spark running on the
EC2 cloud. By exploiting data locality, Kira SE achieves a 2.5x speedup over an
equivalent C program when analyzing a 1TB dataset using 512 cores on the Amazon
EC2 cloud. Furthermore, we show that by leveraging software originally designed
for big data infrastructure, Kira SE achieves competitive performance to the C
implementation running on the NERSC Edison supercomputer. Our experience with
Kira indicates that emerging Big Data platforms such as Apache Spark are a
performant alternative for many-task scientific applications.
| no_new_dataset | 0.939137 |
1602.03418 | Swami Sankaranarayanan | Swami Sankaranarayanan, Azadeh Alavi, Rama Chellappa | Triplet Similarity Embedding for Face Verification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present an unconstrained face verification algorithm and
evaluate it on the recently released IJB-A dataset that aims to push the
boundaries of face verification methods. The proposed algorithm couples a deep
CNN-based approach with a low-dimensional discriminative embedding learnt using
triplet similarity constraints in a large margin fashion. Aside from yielding
performance improvement, this embedding provides significant advantages in
terms of memory and post-processing operations like hashing and visualization.
Experiments on the IJB-A dataset show that the proposed algorithm outperforms
state of the art methods in verification and identification metrics, while
requiring less training time.
| [
{
"version": "v1",
"created": "Wed, 10 Feb 2016 15:48:47 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Mar 2016 18:06:34 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Sankaranarayanan",
"Swami",
""
],
[
"Alavi",
"Azadeh",
""
],
[
"Chellappa",
"Rama",
""
]
] | TITLE: Triplet Similarity Embedding for Face Verification
ABSTRACT: In this work, we present an unconstrained face verification algorithm and
evaluate it on the recently released IJB-A dataset that aims to push the
boundaries of face verification methods. The proposed algorithm couples a deep
CNN-based approach with a low-dimensional discriminative embedding learnt using
triplet similarity constraints in a large margin fashion. Aside from yielding
performance improvement, this embedding provides significant advantages in
terms of memory and post-processing operations like hashing and visualization.
Experiments on the IJB-A dataset show that the proposed algorithm outperforms
state of the art methods in verification and identification metrics, while
requiring less training time.
| no_new_dataset | 0.878991 |
1602.04984 | Hyo-Eun Kim | Hyo-Eun Kim and Sangheum Hwang | Deconvolutional Feature Stacking for Weakly-Supervised Semantic
Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A weakly-supervised semantic segmentation framework with a tied
deconvolutional neural network is presented. Each deconvolution layer in the
framework consists of unpooling and deconvolution operations. 'Unpooling'
upsamples the input feature map based on unpooling switches defined by
corresponding convolution layer's pooling operation. 'Deconvolution' convolves
the input unpooled features by using convolutional weights tied with the
corresponding convolution layer's convolution operation. The
unpooling-deconvolution combination helps to eliminate less discriminative
features in a feature extraction stage, since output features of the
deconvolution layer are reconstructed from the most discriminative unpooled
features instead of the raw one. This results in reduction of false positives
in a pixel-level inference stage. All the feature maps restored from the entire
deconvolution layers can constitute a rich discriminative feature set according
to different abstraction levels. Those features are stacked to be selectively
used for generating class-specific activation maps. Under the weak supervision
(image-level labels), the proposed framework shows promising results on lesion
segmentation in medical images (chest X-rays) and achieves state-of-the-art
performance on the PASCAL VOC segmentation dataset in the same experimental
condition.
| [
{
"version": "v1",
"created": "Tue, 16 Feb 2016 11:05:24 GMT"
},
{
"version": "v2",
"created": "Thu, 18 Feb 2016 07:46:24 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Mar 2016 08:22:30 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Kim",
"Hyo-Eun",
""
],
[
"Hwang",
"Sangheum",
""
]
] | TITLE: Deconvolutional Feature Stacking for Weakly-Supervised Semantic
Segmentation
ABSTRACT: A weakly-supervised semantic segmentation framework with a tied
deconvolutional neural network is presented. Each deconvolution layer in the
framework consists of unpooling and deconvolution operations. 'Unpooling'
upsamples the input feature map based on unpooling switches defined by
corresponding convolution layer's pooling operation. 'Deconvolution' convolves
the input unpooled features by using convolutional weights tied with the
corresponding convolution layer's convolution operation. The
unpooling-deconvolution combination helps to eliminate less discriminative
features in a feature extraction stage, since output features of the
deconvolution layer are reconstructed from the most discriminative unpooled
features instead of the raw one. This results in reduction of false positives
in a pixel-level inference stage. All the feature maps restored from the entire
deconvolution layers can constitute a rich discriminative feature set according
to different abstraction levels. Those features are stacked to be selectively
used for generating class-specific activation maps. Under the weak supervision
(image-level labels), the proposed framework shows promising results on lesion
segmentation in medical images (chest X-rays) and achieves state-of-the-art
performance on the PASCAL VOC segmentation dataset in the same experimental
condition.
| no_new_dataset | 0.950134 |
1603.01979 | Haewoon Kwak | Haewoon Kwak and Jisun An | Two Tales of the World: Comparison of Widely Used World News Datasets
GDELT and EventRegistry | To be appeared in ICWSM'16 | null | null | null | cs.DL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we compare GDELT and Event Registry, which monitor news
articles worldwide and provide big data to researchers regarding scale, news
sources, and news geography. We found significant differences in scale and news
sources, but surprisingly, we observed high similarity in news geography
between the two datasets.
| [
{
"version": "v1",
"created": "Mon, 7 Mar 2016 09:25:57 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Kwak",
"Haewoon",
""
],
[
"An",
"Jisun",
""
]
] | TITLE: Two Tales of the World: Comparison of Widely Used World News Datasets
GDELT and EventRegistry
ABSTRACT: In this work, we compare GDELT and Event Registry, which monitor news
articles worldwide and provide big data to researchers regarding scale, news
sources, and news geography. We found significant differences in scale and news
sources, but surprisingly, we observed high similarity in news geography
between the two datasets.
| no_new_dataset | 0.943764 |
1603.03783 | Sachin Mehta | Sachin Mehta and Balakrishnan Prabhakaran | Region Graph Based Method for Multi-Object Detection and Tracking using
Depth Cameras | Accepted in IEEE Winter Conference in Computer Vision (WACV'16) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a multi-object detection and tracking method using
depth cameras. Depth maps are very noisy and obscure in object detection. We
first propose a region-based method to suppress high magnitude noise which
cannot be filtered using spatial filters. Second, the proposed method detect
Region of Interests by temporal learning which are then tracked using weighted
graph-based approach. We demonstrate the performance of the proposed method on
standard depth camera datasets with and without object occlusions. Experimental
results show that the proposed method is able to suppress high magnitude noise
in depth maps and detect/track the objects (with and without occlusion).
| [
{
"version": "v1",
"created": "Fri, 11 Mar 2016 21:06:35 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Mehta",
"Sachin",
""
],
[
"Prabhakaran",
"Balakrishnan",
""
]
] | TITLE: Region Graph Based Method for Multi-Object Detection and Tracking using
Depth Cameras
ABSTRACT: In this paper, we propose a multi-object detection and tracking method using
depth cameras. Depth maps are very noisy and obscure in object detection. We
first propose a region-based method to suppress high magnitude noise which
cannot be filtered using spatial filters. Second, the proposed method detect
Region of Interests by temporal learning which are then tracked using weighted
graph-based approach. We demonstrate the performance of the proposed method on
standard depth camera datasets with and without object occlusions. Experimental
results show that the proposed method is able to suppress high magnitude noise
in depth maps and detect/track the objects (with and without occlusion).
| no_new_dataset | 0.949856 |
1603.03827 | Franck Dernoncourt | Ji Young Lee, Franck Dernoncourt | Sequential Short-Text Classification with Recurrent and Convolutional
Neural Networks | Accepted as a conference paper at NAACL 2016 | null | null | null | cs.CL cs.AI cs.LG cs.NE stat.ML | http://creativecommons.org/licenses/by/4.0/ | Recent approaches based on artificial neural networks (ANNs) have shown
promising results for short-text classification. However, many short texts
occur in sequences (e.g., sentences in a document or utterances in a dialog),
and most existing ANN-based systems do not leverage the preceding short texts
when classifying a subsequent one. In this work, we present a model based on
recurrent neural networks and convolutional neural networks that incorporates
the preceding short texts. Our model achieves state-of-the-art results on three
different datasets for dialog act prediction.
| [
{
"version": "v1",
"created": "Sat, 12 Mar 2016 00:02:51 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Lee",
"Ji Young",
""
],
[
"Dernoncourt",
"Franck",
""
]
] | TITLE: Sequential Short-Text Classification with Recurrent and Convolutional
Neural Networks
ABSTRACT: Recent approaches based on artificial neural networks (ANNs) have shown
promising results for short-text classification. However, many short texts
occur in sequences (e.g., sentences in a document or utterances in a dialog),
and most existing ANN-based systems do not leverage the preceding short texts
when classifying a subsequent one. In this work, we present a model based on
recurrent neural networks and convolutional neural networks that incorporates
the preceding short texts. Our model achieves state-of-the-art results on three
different datasets for dialog act prediction.
| no_new_dataset | 0.953923 |
1603.03836 | Amirali Aghazadeh | Amirali Aghazadeh, Andrew Lan, Anshumali Shrivastava, Richard Baraniuk | Near-Isometric Binary Hashing for Large-scale Datasets | null | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a scalable algorithm to learn binary hash codes for indexing
large-scale datasets. Near-isometric binary hashing (NIBH) is a data-dependent
hashing scheme that quantizes the output of a learned low-dimensional embedding
to obtain a binary hash code. In contrast to conventional hashing schemes,
which typically rely on an $\ell_2$-norm (i.e., average distortion)
minimization, NIBH is based on a $\ell_{\infty}$-norm (i.e., worst-case
distortion) minimization that provides several benefits, including superior
distance, ranking, and near-neighbor preservation performance. We develop a
practical and efficient algorithm for NIBH based on column generation that
scales well to large datasets. A range of experimental evaluations demonstrate
the superiority of NIBH over ten state-of-the-art binary hashing schemes.
| [
{
"version": "v1",
"created": "Sat, 12 Mar 2016 01:04:50 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Aghazadeh",
"Amirali",
""
],
[
"Lan",
"Andrew",
""
],
[
"Shrivastava",
"Anshumali",
""
],
[
"Baraniuk",
"Richard",
""
]
] | TITLE: Near-Isometric Binary Hashing for Large-scale Datasets
ABSTRACT: We develop a scalable algorithm to learn binary hash codes for indexing
large-scale datasets. Near-isometric binary hashing (NIBH) is a data-dependent
hashing scheme that quantizes the output of a learned low-dimensional embedding
to obtain a binary hash code. In contrast to conventional hashing schemes,
which typically rely on an $\ell_2$-norm (i.e., average distortion)
minimization, NIBH is based on a $\ell_{\infty}$-norm (i.e., worst-case
distortion) minimization that provides several benefits, including superior
distance, ranking, and near-neighbor preservation performance. We develop a
practical and efficient algorithm for NIBH based on column generation that
scales well to large datasets. A range of experimental evaluations demonstrate
the superiority of NIBH over ten state-of-the-art binary hashing schemes.
| no_new_dataset | 0.949435 |
1603.03984 | Rudrasis Chakraborty Mr | Rudrasis Chakraborty, Dohyung Seo and Baba C. Vemuri | An efficient Exact-PGA algorithm for constant curvature manifolds | Accepted in CVPR (IEEE Conference on Computer Vision and Pattern
Recognition) 2016 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Manifold-valued datasets are widely encountered in many computer vision
tasks. A non-linear analog of the PCA, called the Principal Geodesic Analysis
(PGA) suited for data lying on Riemannian manifolds was reported in literature
a decade ago. Since the objective function in PGA is highly non-linear and hard
to solve efficiently in general, researchers have proposed a linear
approximation. Though this linear approximation is easy to compute, it lacks
accuracy especially when the data exhibits a large variance. Recently, an
alternative called exact PGA was proposed which tries to solve the optimization
without any linearization. For general Riemannian manifolds, though it gives
better accuracy than the original (linearized) PGA, for data that exhibit large
variance, the optimization is not computationally efficient. In this paper, we
propose an efficient exact PGA for constant curvature Riemannian manifolds
(CCM-EPGA). CCM-EPGA differs significantly from existing PGA algorithms in two
aspects, (i) the distance between a given manifold-valued data point and the
principal submanifold is computed analytically and thus no optimization is
required as in existing methods. (ii) Unlike the existing PGA algorithms, the
descent into codimension-1 submanifolds does not require any optimization but
is accomplished through the use of the Rimeannian inverse Exponential map and
the parallel transport operations. We present theoretical and experimental
results for constant curvature Riemannian manifolds depicting favorable
performance of CCM-EPGA compared to existing PGA algorithms. We also present
data reconstruction from principal components and directions which has not been
presented in literature in this setting.
| [
{
"version": "v1",
"created": "Sun, 13 Mar 2016 02:57:34 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Chakraborty",
"Rudrasis",
""
],
[
"Seo",
"Dohyung",
""
],
[
"Vemuri",
"Baba C.",
""
]
] | TITLE: An efficient Exact-PGA algorithm for constant curvature manifolds
ABSTRACT: Manifold-valued datasets are widely encountered in many computer vision
tasks. A non-linear analog of the PCA, called the Principal Geodesic Analysis
(PGA) suited for data lying on Riemannian manifolds was reported in literature
a decade ago. Since the objective function in PGA is highly non-linear and hard
to solve efficiently in general, researchers have proposed a linear
approximation. Though this linear approximation is easy to compute, it lacks
accuracy especially when the data exhibits a large variance. Recently, an
alternative called exact PGA was proposed which tries to solve the optimization
without any linearization. For general Riemannian manifolds, though it gives
better accuracy than the original (linearized) PGA, for data that exhibit large
variance, the optimization is not computationally efficient. In this paper, we
propose an efficient exact PGA for constant curvature Riemannian manifolds
(CCM-EPGA). CCM-EPGA differs significantly from existing PGA algorithms in two
aspects, (i) the distance between a given manifold-valued data point and the
principal submanifold is computed analytically and thus no optimization is
required as in existing methods. (ii) Unlike the existing PGA algorithms, the
descent into codimension-1 submanifolds does not require any optimization but
is accomplished through the use of the Rimeannian inverse Exponential map and
the parallel transport operations. We present theoretical and experimental
results for constant curvature Riemannian manifolds depicting favorable
performance of CCM-EPGA compared to existing PGA algorithms. We also present
data reconstruction from principal components and directions which has not been
presented in literature in this setting.
| no_new_dataset | 0.952131 |
1603.04002 | Richard Nock | Giorgio Patrini, Richard Nock, Stephen Hardy, Tiberio Caetano | Fast Learning from Distributed Datasets without Entity Matching | null | null | null | null | cs.LG cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Consider the following data fusion scenario: two datasets/peers contain the
same real-world entities described using partially shared features, e.g.
banking and insurance company records of the same customer base. Our goal is to
learn a classifier in the cross product space of the two domains, in the hard
case in which no shared ID is available -- e.g. due to anonymization.
Traditionally, the problem is approached by first addressing entity matching
and subsequently learning the classifier in a standard manner. We present an
end-to-end solution which bypasses matching entities, based on the recently
introduced concept of Rademacher observations (rados). Informally, we replace
the minimisation of a loss over examples, which requires to solve entity
resolution, by the equivalent minimisation of a (different) loss over rados.
Among others, key properties we show are (i) a potentially huge subset of these
rados does not require to perform entity matching, and (ii) the algorithm that
provably minimizes the rado loss over these rados has time and space
complexities smaller than the algorithm minimizing the equivalent example loss.
Last, we relax a key assumption of the model, that the data is vertically
partitioned among peers --- in this case, we would not even know the existence
of a solution to entity resolution. In this more general setting, experiments
validate the possibility of significantly beating even the optimal peer in
hindsight.
| [
{
"version": "v1",
"created": "Sun, 13 Mar 2016 06:03:39 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Patrini",
"Giorgio",
""
],
[
"Nock",
"Richard",
""
],
[
"Hardy",
"Stephen",
""
],
[
"Caetano",
"Tiberio",
""
]
] | TITLE: Fast Learning from Distributed Datasets without Entity Matching
ABSTRACT: Consider the following data fusion scenario: two datasets/peers contain the
same real-world entities described using partially shared features, e.g.
banking and insurance company records of the same customer base. Our goal is to
learn a classifier in the cross product space of the two domains, in the hard
case in which no shared ID is available -- e.g. due to anonymization.
Traditionally, the problem is approached by first addressing entity matching
and subsequently learning the classifier in a standard manner. We present an
end-to-end solution which bypasses matching entities, based on the recently
introduced concept of Rademacher observations (rados). Informally, we replace
the minimisation of a loss over examples, which requires to solve entity
resolution, by the equivalent minimisation of a (different) loss over rados.
Among others, key properties we show are (i) a potentially huge subset of these
rados does not require to perform entity matching, and (ii) the algorithm that
provably minimizes the rado loss over these rados has time and space
complexities smaller than the algorithm minimizing the equivalent example loss.
Last, we relax a key assumption of the model, that the data is vertically
partitioned among peers --- in this case, we would not even know the existence
of a solution to entity resolution. In this more general setting, experiments
validate the possibility of significantly beating even the optimal peer in
hindsight.
| no_new_dataset | 0.948489 |
1603.04010 | Sofiane Abbar | Sofiane Abbar, Tahar Zanouda, Laure Berti-Equille, Javier
Borge-Holthoefer | Using Twitter to Understand Public Interest in Climate Change: The case
of Qatar | Will appear in the proceedings of the International Workshop on
Social Media for Environment and Ecological Monitoring (SWEEM'16) | null | null | null | cs.SI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Climate change has received an extensive attention from public opinion in the
last couple of years, after being considered for decades as an exclusive
scientific debate. Governments and world-wide organizations such as the United
Nations are working more than ever on raising and maintaining public awareness
toward this global issue. In the present study, we examine and analyze Climate
Change conversations in Qatar's Twittersphere, and sense public awareness
towards this global and shared problem in general, and its various related
topics in particular. Such topics include but are not limited to politics,
economy, disasters, energy and sandstorms. To address this concern, we collect
and analyze a large dataset of 109 million tweets posted by 98K distinct users
living in Qatar -- one of the largest emitters of CO2 worldwide. We use a
taxonomy of climate change topics created as part of the United Nations Pulse
project to capture the climate change discourse in more than 36K tweets. We
also examine which topics people refer to when they discuss climate change, and
perform different analysis to understand the temporal dynamics of public
interest toward these topics.
| [
{
"version": "v1",
"created": "Sun, 13 Mar 2016 08:39:19 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Abbar",
"Sofiane",
""
],
[
"Zanouda",
"Tahar",
""
],
[
"Berti-Equille",
"Laure",
""
],
[
"Borge-Holthoefer",
"Javier",
""
]
] | TITLE: Using Twitter to Understand Public Interest in Climate Change: The case
of Qatar
ABSTRACT: Climate change has received an extensive attention from public opinion in the
last couple of years, after being considered for decades as an exclusive
scientific debate. Governments and world-wide organizations such as the United
Nations are working more than ever on raising and maintaining public awareness
toward this global issue. In the present study, we examine and analyze Climate
Change conversations in Qatar's Twittersphere, and sense public awareness
towards this global and shared problem in general, and its various related
topics in particular. Such topics include but are not limited to politics,
economy, disasters, energy and sandstorms. To address this concern, we collect
and analyze a large dataset of 109 million tweets posted by 98K distinct users
living in Qatar -- one of the largest emitters of CO2 worldwide. We use a
taxonomy of climate change topics created as part of the United Nations Pulse
project to capture the climate change discourse in more than 36K tweets. We
also examine which topics people refer to when they discuss climate change, and
perform different analysis to understand the temporal dynamics of public
interest toward these topics.
| new_dataset | 0.845113 |
1603.04026 | Huamin Ren | Huamin Ren, Hong Pan, S{\o}ren Ingvor Olsen, Thomas B. Moeslund | A comprehensive study of sparse codes on abnormality detection | 7 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sparse representation has been applied successfully in abnormal event
detection, in which the baseline is to learn a dictionary accompanied by sparse
codes. While much emphasis is put on discriminative dictionary construction,
there are no comparative studies of sparse codes regarding abnormality
detection. We comprehensively study two types of sparse codes solutions -
greedy algorithms and convex L1-norm solutions - and their impact on
abnormality detection performance. We also propose our framework of combining
sparse codes with different detection methods. Our comparative experiments are
carried out from various angles to better understand the applicability of
sparse codes, including computation time, reconstruction error, sparsity,
detection accuracy, and their performance combining various detection methods.
Experiments show that combining OMP codes with maximum coordinate detection
could achieve state-of-the-art performance on the UCSD dataset [14].
| [
{
"version": "v1",
"created": "Sun, 13 Mar 2016 13:13:10 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Ren",
"Huamin",
""
],
[
"Pan",
"Hong",
""
],
[
"Olsen",
"Søren Ingvor",
""
],
[
"Moeslund",
"Thomas B.",
""
]
] | TITLE: A comprehensive study of sparse codes on abnormality detection
ABSTRACT: Sparse representation has been applied successfully in abnormal event
detection, in which the baseline is to learn a dictionary accompanied by sparse
codes. While much emphasis is put on discriminative dictionary construction,
there are no comparative studies of sparse codes regarding abnormality
detection. We comprehensively study two types of sparse codes solutions -
greedy algorithms and convex L1-norm solutions - and their impact on
abnormality detection performance. We also propose our framework of combining
sparse codes with different detection methods. Our comparative experiments are
carried out from various angles to better understand the applicability of
sparse codes, including computation time, reconstruction error, sparsity,
detection accuracy, and their performance combining various detection methods.
Experiments show that combining OMP codes with maximum coordinate detection
could achieve state-of-the-art performance on the UCSD dataset [14].
| no_new_dataset | 0.940898 |
1603.04042 | Ning Xu | Ning Xu, Brian Price, Scott Cohen, Jimei Yang, Thomas Huang | Deep Interactive Object Selection | Computer Vision and Pattern Recognition | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Interactive object selection is a very important research problem and has
many applications. Previous algorithms require substantial user interactions to
estimate the foreground and background distributions. In this paper, we present
a novel deep learning based algorithm which has a much better understanding of
objectness and thus can reduce user interactions to just a few clicks. Our
algorithm transforms user provided positive and negative clicks into two
Euclidean distance maps which are then concatenated with the RGB channels of
images to compose (image, user interactions) pairs. We generate many of such
pairs by combining several random sampling strategies to model user click
patterns and use them to fine tune deep Fully Convolutional Networks (FCNs).
Finally the output probability maps of our FCN 8s model is integrated with
graph cut optimization to refine the boundary segments. Our model is trained on
the PASCAL segmentation dataset and evaluated on other datasets with different
object classes. Experimental results on both seen and unseen objects clearly
demonstrate that our algorithm has a good generalization ability and is
superior to all existing interactive object selection approaches.
| [
{
"version": "v1",
"created": "Sun, 13 Mar 2016 15:42:34 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Xu",
"Ning",
""
],
[
"Price",
"Brian",
""
],
[
"Cohen",
"Scott",
""
],
[
"Yang",
"Jimei",
""
],
[
"Huang",
"Thomas",
""
]
] | TITLE: Deep Interactive Object Selection
ABSTRACT: Interactive object selection is a very important research problem and has
many applications. Previous algorithms require substantial user interactions to
estimate the foreground and background distributions. In this paper, we present
a novel deep learning based algorithm which has a much better understanding of
objectness and thus can reduce user interactions to just a few clicks. Our
algorithm transforms user provided positive and negative clicks into two
Euclidean distance maps which are then concatenated with the RGB channels of
images to compose (image, user interactions) pairs. We generate many of such
pairs by combining several random sampling strategies to model user click
patterns and use them to fine tune deep Fully Convolutional Networks (FCNs).
Finally the output probability maps of our FCN 8s model is integrated with
graph cut optimization to refine the boundary segments. Our model is trained on
the PASCAL segmentation dataset and evaluated on other datasets with different
object classes. Experimental results on both seen and unseen objects clearly
demonstrate that our algorithm has a good generalization ability and is
superior to all existing interactive object selection approaches.
| no_new_dataset | 0.950041 |
1603.04265 | Tae Hyun Kim | Tae Hyun Kim, Seungjun Nah, and Kyoung Mu Lee | Dynamic Scene Deblurring using a Locally Adaptive Linear Blur Model | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | State-of-the-art video deblurring methods cannot handle blurry videos
recorded in dynamic scenes, since they are built under a strong assumption that
the captured scenes are static. Contrary to the existing methods, we propose a
video deblurring algorithm that can deal with general blurs inherent in dynamic
scenes. To handle general and locally varying blurs caused by various sources,
such as moving objects, camera shake, depth variation, and defocus, we estimate
pixel-wise non-uniform blur kernels. We infer bidirectional optical flows to
handle motion blurs, and also estimate Gaussian blur maps to remove optical
blur from defocus in our new blur model. Therefore, we propose a single energy
model that jointly estimates optical flows, defocus blur maps and latent
frames. We also provide a framework and efficient solvers to minimize the
proposed energy model. By optimizing the energy model, we achieve significant
improvements in removing general blurs, estimating optical flows, and extending
depth-of-field in blurry frames. Moreover, in this work, to evaluate the
performance of non-uniform deblurring methods objectively, we have constructed
a new realistic dataset with ground truths. In addition, extensive experimental
on publicly available challenging video data demonstrate that the proposed
method produces qualitatively superior performance than the state-of-the-art
methods which often fail in either deblurring or optical flow estimation.
| [
{
"version": "v1",
"created": "Mon, 14 Mar 2016 13:57:19 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Kim",
"Tae Hyun",
""
],
[
"Nah",
"Seungjun",
""
],
[
"Lee",
"Kyoung Mu",
""
]
] | TITLE: Dynamic Scene Deblurring using a Locally Adaptive Linear Blur Model
ABSTRACT: State-of-the-art video deblurring methods cannot handle blurry videos
recorded in dynamic scenes, since they are built under a strong assumption that
the captured scenes are static. Contrary to the existing methods, we propose a
video deblurring algorithm that can deal with general blurs inherent in dynamic
scenes. To handle general and locally varying blurs caused by various sources,
such as moving objects, camera shake, depth variation, and defocus, we estimate
pixel-wise non-uniform blur kernels. We infer bidirectional optical flows to
handle motion blurs, and also estimate Gaussian blur maps to remove optical
blur from defocus in our new blur model. Therefore, we propose a single energy
model that jointly estimates optical flows, defocus blur maps and latent
frames. We also provide a framework and efficient solvers to minimize the
proposed energy model. By optimizing the energy model, we achieve significant
improvements in removing general blurs, estimating optical flows, and extending
depth-of-field in blurry frames. Moreover, in this work, to evaluate the
performance of non-uniform deblurring methods objectively, we have constructed
a new realistic dataset with ground truths. In addition, extensive experimental
on publicly available challenging video data demonstrate that the proposed
method produces qualitatively superior performance than the state-of-the-art
methods which often fail in either deblurring or optical flow estimation.
| new_dataset | 0.955899 |
1603.04319 | Jalal Etesami | Jalal Etesami, Negar Kiyavash, Kun Zhang, Kushagra Singhal | Learning Network of Multivariate Hawkes Processes: A Time Series
Approach | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning the influence structure of multiple time series data is of great
interest to many disciplines. This paper studies the problem of recovering the
causal structure in network of multivariate linear Hawkes processes. In such
processes, the occurrence of an event in one process affects the probability of
occurrence of new events in some other processes. Thus, a natural notion of
causality exists between such processes captured by the support of the
excitation matrix. We show that the resulting causal influence network is
equivalent to the Directed Information graph (DIG) of the processes, which
encodes the causal factorization of the joint distribution of the processes.
Furthermore, we present an algorithm for learning the support of excitation
matrix (or equivalently the DIG). The performance of the algorithm is evaluated
on synthesized multivariate Hawkes networks as well as a stock market and
MemeTracker real-world dataset.
| [
{
"version": "v1",
"created": "Mon, 14 Mar 2016 16:08:26 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Etesami",
"Jalal",
""
],
[
"Kiyavash",
"Negar",
""
],
[
"Zhang",
"Kun",
""
],
[
"Singhal",
"Kushagra",
""
]
] | TITLE: Learning Network of Multivariate Hawkes Processes: A Time Series
Approach
ABSTRACT: Learning the influence structure of multiple time series data is of great
interest to many disciplines. This paper studies the problem of recovering the
causal structure in network of multivariate linear Hawkes processes. In such
processes, the occurrence of an event in one process affects the probability of
occurrence of new events in some other processes. Thus, a natural notion of
causality exists between such processes captured by the support of the
excitation matrix. We show that the resulting causal influence network is
equivalent to the Directed Information graph (DIG) of the processes, which
encodes the causal factorization of the joint distribution of the processes.
Furthermore, we present an algorithm for learning the support of excitation
matrix (or equivalently the DIG). The performance of the algorithm is evaluated
on synthesized multivariate Hawkes networks as well as a stock market and
MemeTracker real-world dataset.
| no_new_dataset | 0.949809 |
1603.04327 | Ibrahim Sadek | Ibrahim Sadek | Automatic Discrimination of Color Retinal Images using the Bag of Words
Approach | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diabetic retinopathy (DR) and age related macular degeneration (ARMD) are
among the major causes of visual impairment worldwide. DR is mainly
characterized by red spots, namely microaneurysms and bright lesions,
specifically exudates whereas ARMD is mainly identified by tiny yellow or white
deposits called drusen. Since exudates might be the only manifestation of the
early diabetic retinopathy, there is an increase demand for automatic
retinopathy diagnosis. Exudates and drusen may share similar appearances, thus
discriminating between them is of interest to enhance screening performance. In
this research, we investigative the role of bag of words approach in the
automatic diagnosis of retinopathy diabetes. We proposed to use a single based
and multiple based methods for the construction of the visual dictionary by
combining the histogram of word occurrences from each dictionary and building a
single histogram. The introduced approach is evaluated for automatic diagnosis
of normal and abnormal color fundus images with bright lesions. This approach
has been implemented on 430 fundus images, including six publicly available
datasets, in addition to one local dataset. The mean accuracies reported are
97.2% and 99.77% for single based and multiple based dictionaries respectively.
| [
{
"version": "v1",
"created": "Mon, 14 Mar 2016 16:26:32 GMT"
}
] | 2016-03-15T00:00:00 | [
[
"Sadek",
"Ibrahim",
""
]
] | TITLE: Automatic Discrimination of Color Retinal Images using the Bag of Words
Approach
ABSTRACT: Diabetic retinopathy (DR) and age related macular degeneration (ARMD) are
among the major causes of visual impairment worldwide. DR is mainly
characterized by red spots, namely microaneurysms and bright lesions,
specifically exudates whereas ARMD is mainly identified by tiny yellow or white
deposits called drusen. Since exudates might be the only manifestation of the
early diabetic retinopathy, there is an increase demand for automatic
retinopathy diagnosis. Exudates and drusen may share similar appearances, thus
discriminating between them is of interest to enhance screening performance. In
this research, we investigative the role of bag of words approach in the
automatic diagnosis of retinopathy diabetes. We proposed to use a single based
and multiple based methods for the construction of the visual dictionary by
combining the histogram of word occurrences from each dictionary and building a
single histogram. The introduced approach is evaluated for automatic diagnosis
of normal and abnormal color fundus images with bright lesions. This approach
has been implemented on 430 fundus images, including six publicly available
datasets, in addition to one local dataset. The mean accuracies reported are
97.2% and 99.77% for single based and multiple based dictionaries respectively.
| no_new_dataset | 0.883286 |
1512.02895 | Shaoting Zhang | Xiaofan Zhang and Feng Zhou and Yuanqing Lin and Shaoting Zhang | Embedding Label Structures for Fine-Grained Feature Representation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recent algorithms in convolutional neural networks (CNN) considerably advance
the fine-grained image classification, which aims to differentiate subtle
differences among subordinate classes. However, previous studies have rarely
focused on learning a fined-grained and structured feature representation that
is able to locate similar images at different levels of relevance, e.g.,
discovering cars from the same make or the same model, both of which require
high precision. In this paper, we propose two main contributions to tackle this
problem. 1) A multi-task learning framework is designed to effectively learn
fine-grained feature representations by jointly optimizing both classification
and similarity constraints. 2) To model the multi-level relevance, label
structures such as hierarchy or shared attributes are seamlessly embedded into
the framework by generalizing the triplet loss. Extensive and thorough
experiments have been conducted on three fine-grained datasets, i.e., the
Stanford car, the car-333, and the food datasets, which contain either
hierarchical labels or shared attributes. Our proposed method has achieved very
competitive performance, i.e., among state-of-the-art classification accuracy.
More importantly, it significantly outperforms previous fine-grained feature
representations for image retrieval at different levels of relevance.
| [
{
"version": "v1",
"created": "Wed, 9 Dec 2015 15:22:26 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Mar 2016 02:59:36 GMT"
}
] | 2016-03-14T00:00:00 | [
[
"Zhang",
"Xiaofan",
""
],
[
"Zhou",
"Feng",
""
],
[
"Lin",
"Yuanqing",
""
],
[
"Zhang",
"Shaoting",
""
]
] | TITLE: Embedding Label Structures for Fine-Grained Feature Representation
ABSTRACT: Recent algorithms in convolutional neural networks (CNN) considerably advance
the fine-grained image classification, which aims to differentiate subtle
differences among subordinate classes. However, previous studies have rarely
focused on learning a fined-grained and structured feature representation that
is able to locate similar images at different levels of relevance, e.g.,
discovering cars from the same make or the same model, both of which require
high precision. In this paper, we propose two main contributions to tackle this
problem. 1) A multi-task learning framework is designed to effectively learn
fine-grained feature representations by jointly optimizing both classification
and similarity constraints. 2) To model the multi-level relevance, label
structures such as hierarchy or shared attributes are seamlessly embedded into
the framework by generalizing the triplet loss. Extensive and thorough
experiments have been conducted on three fine-grained datasets, i.e., the
Stanford car, the car-333, and the food datasets, which contain either
hierarchical labels or shared attributes. Our proposed method has achieved very
competitive performance, i.e., among state-of-the-art classification accuracy.
More importantly, it significantly outperforms previous fine-grained feature
representations for image retrieval at different levels of relevance.
| no_new_dataset | 0.947137 |
1603.03541 | Chenxia Wu | Chenxia Wu, Jiemi Zhang, Ozan Sener, Bart Selman, Silvio Savarese,
Ashutosh Saxena | Watch-n-Patch: Unsupervised Learning of Actions and Relations | arXiv admin note: text overlap with arXiv:1512.04208 | null | null | null | cs.CV cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There is a large variation in the activities that humans perform in their
everyday lives. We consider modeling these composite human activities which
comprises multiple basic level actions in a completely unsupervised setting.
Our model learns high-level co-occurrence and temporal relations between the
actions. We consider the video as a sequence of short-term action clips, which
contains human-words and object-words. An activity is about a set of
action-topics and object-topics indicating which actions are present and which
objects are interacting with. We then propose a new probabilistic model
relating the words and the topics. It allows us to model long-range action
relations that commonly exist in the composite activities, which is challenging
in previous works. We apply our model to the unsupervised action segmentation
and clustering, and to a novel application that detects forgotten actions,
which we call action patching. For evaluation, we contribute a new challenging
RGB-D activity video dataset recorded by the new Kinect v2, which contains
several human daily activities as compositions of multiple actions interacting
with different objects. Moreover, we develop a robotic system that watches
people and reminds people by applying our action patching algorithm. Our
robotic setup can be easily deployed on any assistive robot.
| [
{
"version": "v1",
"created": "Fri, 11 Mar 2016 07:13:59 GMT"
}
] | 2016-03-14T00:00:00 | [
[
"Wu",
"Chenxia",
""
],
[
"Zhang",
"Jiemi",
""
],
[
"Sener",
"Ozan",
""
],
[
"Selman",
"Bart",
""
],
[
"Savarese",
"Silvio",
""
],
[
"Saxena",
"Ashutosh",
""
]
] | TITLE: Watch-n-Patch: Unsupervised Learning of Actions and Relations
ABSTRACT: There is a large variation in the activities that humans perform in their
everyday lives. We consider modeling these composite human activities which
comprises multiple basic level actions in a completely unsupervised setting.
Our model learns high-level co-occurrence and temporal relations between the
actions. We consider the video as a sequence of short-term action clips, which
contains human-words and object-words. An activity is about a set of
action-topics and object-topics indicating which actions are present and which
objects are interacting with. We then propose a new probabilistic model
relating the words and the topics. It allows us to model long-range action
relations that commonly exist in the composite activities, which is challenging
in previous works. We apply our model to the unsupervised action segmentation
and clustering, and to a novel application that detects forgotten actions,
which we call action patching. For evaluation, we contribute a new challenging
RGB-D activity video dataset recorded by the new Kinect v2, which contains
several human daily activities as compositions of multiple actions interacting
with different objects. Moreover, we develop a robotic system that watches
people and reminds people by applying our action patching algorithm. Our
robotic setup can be easily deployed on any assistive robot.
| new_dataset | 0.955026 |
1603.03627 | Roberto Luis Shinmoto Torres | Roberto L. Shinmoto Torres and Damith C. Ranasinghe and Qinfeng Shi
and Anton van den Hengel | Learning from Imbalanced Multiclass Sequential Data Streams Using
Dynamically Weighted Conditional Random Fields | 28 pages, 8 figures, 1 table | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The present study introduces a method for improving the classification
performance of imbalanced multiclass data streams from wireless body worn
sensors. Data imbalance is an inherent problem in activity recognition caused
by the irregular time distribution of activities, which are sequential and
dependent on previous movements. We use conditional random fields (CRF), a
graphical model for structured classification, to take advantage of
dependencies between activities in a sequence. However, CRFs do not consider
the negative effects of class imbalance during training. We propose a
class-wise dynamically weighted CRF (dWCRF) where weights are automatically
determined during training by maximizing the expected overall F-score. Our
results based on three case studies from a healthcare application using a
batteryless body worn sensor, demonstrate that our method, in general, improves
overall and minority class F-score when compared to other CRF based classifiers
and achieves similar or better overall and class-wise performance when compared
to SVM based classifiers under conditions of limited training data. We also
confirm the performance of our approach using an additional battery powered
body worn sensor dataset, achieving similar results in cases of high class
imbalance.
| [
{
"version": "v1",
"created": "Fri, 11 Mar 2016 13:51:37 GMT"
}
] | 2016-03-14T00:00:00 | [
[
"Torres",
"Roberto L. Shinmoto",
""
],
[
"Ranasinghe",
"Damith C.",
""
],
[
"Shi",
"Qinfeng",
""
],
[
"Hengel",
"Anton van den",
""
]
] | TITLE: Learning from Imbalanced Multiclass Sequential Data Streams Using
Dynamically Weighted Conditional Random Fields
ABSTRACT: The present study introduces a method for improving the classification
performance of imbalanced multiclass data streams from wireless body worn
sensors. Data imbalance is an inherent problem in activity recognition caused
by the irregular time distribution of activities, which are sequential and
dependent on previous movements. We use conditional random fields (CRF), a
graphical model for structured classification, to take advantage of
dependencies between activities in a sequence. However, CRFs do not consider
the negative effects of class imbalance during training. We propose a
class-wise dynamically weighted CRF (dWCRF) where weights are automatically
determined during training by maximizing the expected overall F-score. Our
results based on three case studies from a healthcare application using a
batteryless body worn sensor, demonstrate that our method, in general, improves
overall and minority class F-score when compared to other CRF based classifiers
and achieves similar or better overall and class-wise performance when compared
to SVM based classifiers under conditions of limited training data. We also
confirm the performance of our approach using an additional battery powered
body worn sensor dataset, achieving similar results in cases of high class
imbalance.
| no_new_dataset | 0.949995 |
1603.03724 | Niharika Gauraha Niharika Gauraha | Niharika Gauraha, Swapan K. Parui | Efficient Clustering of Correlated Variables and Variable Selection in
High-Dimensional Linear Models | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we introduce Adaptive Cluster Lasso(ACL) method for variable
selection in high dimensional sparse regression models with strongly correlated
variables. To handle correlated variables, the concept of clustering or
grouping variables and then pursuing model fitting is widely accepted. When the
dimension is very high, finding an appropriate group structure is as difficult
as the original problem. The ACL is a three-stage procedure where, at the first
stage, we use the Lasso(or its adaptive or thresholded version) to do initial
selection, then we also include those variables which are not selected by the
Lasso but are strongly correlated with the variables selected by the Lasso. At
the second stage we cluster the variables based on the reduced set of
predictors and in the third stage we perform sparse estimation such as Lasso on
cluster representatives or the group Lasso based on the structures generated by
clustering procedure. We show that our procedure is consistent and efficient in
finding true underlying population group structure(under assumption of
irrepresentable and beta-min conditions). We also study the group selection
consistency of our method and we support the theory using simulated and
pseudo-real dataset examples.
| [
{
"version": "v1",
"created": "Fri, 11 Mar 2016 19:06:33 GMT"
}
] | 2016-03-14T00:00:00 | [
[
"Gauraha",
"Niharika",
""
],
[
"Parui",
"Swapan K.",
""
]
] | TITLE: Efficient Clustering of Correlated Variables and Variable Selection in
High-Dimensional Linear Models
ABSTRACT: In this paper, we introduce Adaptive Cluster Lasso(ACL) method for variable
selection in high dimensional sparse regression models with strongly correlated
variables. To handle correlated variables, the concept of clustering or
grouping variables and then pursuing model fitting is widely accepted. When the
dimension is very high, finding an appropriate group structure is as difficult
as the original problem. The ACL is a three-stage procedure where, at the first
stage, we use the Lasso(or its adaptive or thresholded version) to do initial
selection, then we also include those variables which are not selected by the
Lasso but are strongly correlated with the variables selected by the Lasso. At
the second stage we cluster the variables based on the reduced set of
predictors and in the third stage we perform sparse estimation such as Lasso on
cluster representatives or the group Lasso based on the structures generated by
clustering procedure. We show that our procedure is consistent and efficient in
finding true underlying population group structure(under assumption of
irrepresentable and beta-min conditions). We also study the group selection
consistency of our method and we support the theory using simulated and
pseudo-real dataset examples.
| no_new_dataset | 0.951594 |
1406.5370 | Alexandre d'Aspremont | Fajwel Fogel, Alexandre d'Aspremont, Milan Vojnovic | Spectral Ranking using Seriation | Substantially revised. Accepted by JMLR | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a seriation algorithm for ranking a set of items given pairwise
comparisons between these items. Intuitively, the algorithm assigns similar
rankings to items that compare similarly with all others. It does so by
constructing a similarity matrix from pairwise comparisons, using seriation
methods to reorder this matrix and construct a ranking. We first show that this
spectral seriation algorithm recovers the true ranking when all pairwise
comparisons are observed and consistent with a total order. We then show that
ranking reconstruction is still exact when some pairwise comparisons are
corrupted or missing, and that seriation based spectral ranking is more robust
to noise than classical scoring methods. Finally, we bound the ranking error
when only a random subset of the comparions are observed. An additional benefit
of the seriation formulation is that it allows us to solve semi-supervised
ranking problems. Experiments on both synthetic and real datasets demonstrate
that seriation based spectral ranking achieves competitive and in some cases
superior performance compared to classical ranking methods.
| [
{
"version": "v1",
"created": "Fri, 20 Jun 2014 12:58:46 GMT"
},
{
"version": "v2",
"created": "Thu, 25 Jun 2015 10:21:50 GMT"
},
{
"version": "v3",
"created": "Mon, 18 Jan 2016 18:09:07 GMT"
},
{
"version": "v4",
"created": "Thu, 10 Mar 2016 18:15:19 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Fogel",
"Fajwel",
""
],
[
"d'Aspremont",
"Alexandre",
""
],
[
"Vojnovic",
"Milan",
""
]
] | TITLE: Spectral Ranking using Seriation
ABSTRACT: We describe a seriation algorithm for ranking a set of items given pairwise
comparisons between these items. Intuitively, the algorithm assigns similar
rankings to items that compare similarly with all others. It does so by
constructing a similarity matrix from pairwise comparisons, using seriation
methods to reorder this matrix and construct a ranking. We first show that this
spectral seriation algorithm recovers the true ranking when all pairwise
comparisons are observed and consistent with a total order. We then show that
ranking reconstruction is still exact when some pairwise comparisons are
corrupted or missing, and that seriation based spectral ranking is more robust
to noise than classical scoring methods. Finally, we bound the ranking error
when only a random subset of the comparions are observed. An additional benefit
of the seriation formulation is that it allows us to solve semi-supervised
ranking problems. Experiments on both synthetic and real datasets demonstrate
that seriation based spectral ranking achieves competitive and in some cases
superior performance compared to classical ranking methods.
| no_new_dataset | 0.946498 |
1410.6751 | Christoph Riedl | Christoph Riedl, Richard Zanibbi, Marti A. Hearst, Siyu Zhu, Michael
Menietti, Jason Crusan, Ivan Metelsky, Karim R. Lakhani | Detecting Figures and Part Labels in Patents: Competition-Based
Development of Image Processing Algorithms | null | null | 10.1007/s10032-016-0260-8 | null | cs.CV cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report the findings of a month-long online competition in which
participants developed algorithms for augmenting the digital version of patent
documents published by the United States Patent and Trademark Office (USPTO).
The goal was to detect figures and part labels in U.S. patent drawing pages.
The challenge drew 232 teams of two, of which 70 teams (30%) submitted
solutions. Collectively, teams submitted 1,797 solutions that were compiled on
the competition servers. Participants reported spending an average of 63 hours
developing their solutions, resulting in a total of 5,591 hours of development
time. A manually labeled dataset of 306 patents was used for training, online
system tests, and evaluation. The design and performance of the top-5 systems
are presented, along with a system developed after the competition which
illustrates that winning teams produced near state-of-the-art results under
strict time and computation constraints. For the 1st place system, the harmonic
mean of recall and precision (f-measure) was 88.57% for figure region
detection, 78.81% for figure regions with correctly recognized figure titles,
and 70.98% for part label detection and character recognition. Data and
software from the competition are available through the online UCI Machine
Learning repository to inspire follow-on work by the image processing
community.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 17:45:36 GMT"
},
{
"version": "v2",
"created": "Mon, 27 Oct 2014 10:54:17 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Nov 2014 14:33:11 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Riedl",
"Christoph",
""
],
[
"Zanibbi",
"Richard",
""
],
[
"Hearst",
"Marti A.",
""
],
[
"Zhu",
"Siyu",
""
],
[
"Menietti",
"Michael",
""
],
[
"Crusan",
"Jason",
""
],
[
"Metelsky",
"Ivan",
""
],
[
"Lakhani",
"Karim R.",
""
]
] | TITLE: Detecting Figures and Part Labels in Patents: Competition-Based
Development of Image Processing Algorithms
ABSTRACT: We report the findings of a month-long online competition in which
participants developed algorithms for augmenting the digital version of patent
documents published by the United States Patent and Trademark Office (USPTO).
The goal was to detect figures and part labels in U.S. patent drawing pages.
The challenge drew 232 teams of two, of which 70 teams (30%) submitted
solutions. Collectively, teams submitted 1,797 solutions that were compiled on
the competition servers. Participants reported spending an average of 63 hours
developing their solutions, resulting in a total of 5,591 hours of development
time. A manually labeled dataset of 306 patents was used for training, online
system tests, and evaluation. The design and performance of the top-5 systems
are presented, along with a system developed after the competition which
illustrates that winning teams produced near state-of-the-art results under
strict time and computation constraints. For the 1st place system, the harmonic
mean of recall and precision (f-measure) was 88.57% for figure region
detection, 78.81% for figure regions with correctly recognized figure titles,
and 70.98% for part label detection and character recognition. Data and
software from the competition are available through the online UCI Machine
Learning repository to inspire follow-on work by the image processing
community.
| new_dataset | 0.938576 |
1510.07867 | Rasmus Rothe | Rasmus Rothe and Radu Timofte and Luc Van Gool | Some like it hot - visual guidance for preference prediction | accepted for publication at CVPR 2016 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For people first impressions of someone are of determining importance. They
are hard to alter through further information. This begs the question if a
computer can reach the same judgement. Earlier research has already pointed out
that age, gender, and average attractiveness can be estimated with reasonable
precision. We improve the state-of-the-art, but also predict - based on
someone's known preferences - how much that particular person is attracted to a
novel face. Our computational pipeline comprises a face detector, convolutional
neural networks for the extraction of deep features, standard support vector
regression for gender, age and facial beauty, and - as the main novelties -
visual regularized collaborative filtering to infer inter-person preferences as
well as a novel regression technique for handling visual queries without rating
history. We validate the method using a very large dataset from a dating site
as well as images from celebrities. Our experiments yield convincing results,
i.e. we predict 76% of the ratings correctly solely based on an image, and
reveal some sociologically relevant conclusions. We also validate our
collaborative filtering solution on the standard MovieLens rating dataset,
augmented with movie posters, to predict an individual's movie rating. We
demonstrate our algorithms on howhot.io which went viral around the Internet
with more than 50 million pictures evaluated in the first month.
| [
{
"version": "v1",
"created": "Tue, 27 Oct 2015 11:17:46 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Mar 2016 15:02:19 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Rothe",
"Rasmus",
""
],
[
"Timofte",
"Radu",
""
],
[
"Van Gool",
"Luc",
""
]
] | TITLE: Some like it hot - visual guidance for preference prediction
ABSTRACT: For people first impressions of someone are of determining importance. They
are hard to alter through further information. This begs the question if a
computer can reach the same judgement. Earlier research has already pointed out
that age, gender, and average attractiveness can be estimated with reasonable
precision. We improve the state-of-the-art, but also predict - based on
someone's known preferences - how much that particular person is attracted to a
novel face. Our computational pipeline comprises a face detector, convolutional
neural networks for the extraction of deep features, standard support vector
regression for gender, age and facial beauty, and - as the main novelties -
visual regularized collaborative filtering to infer inter-person preferences as
well as a novel regression technique for handling visual queries without rating
history. We validate the method using a very large dataset from a dating site
as well as images from celebrities. Our experiments yield convincing results,
i.e. we predict 76% of the ratings correctly solely based on an image, and
reveal some sociologically relevant conclusions. We also validate our
collaborative filtering solution on the standard MovieLens rating dataset,
augmented with movie posters, to predict an individual's movie rating. We
demonstrate our algorithms on howhot.io which went viral around the Internet
with more than 50 million pictures evaluated in the first month.
| no_new_dataset | 0.934275 |
1511.04317 | Mansour Ahmadi | Mansour Ahmadi, Dmitry Ulyanov, Stanislav Semenov, Mikhail Trofimov,
Giorgio Giacinto | Novel Feature Extraction, Selection and Fusion for Effective Malware
Family Classification | null | null | null | null | cs.CR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern malware is designed with mutation characteristics, namely polymorphism
and metamorphism, which causes an enormous growth in the number of variants of
malware samples. Categorization of malware samples on the basis of their
behaviors is essential for the computer security community, because they
receive huge number of malware everyday, and the signature extraction process
is usually based on malicious parts characterizing malware families. Microsoft
released a malware classification challenge in 2015 with a huge dataset of near
0.5 terabytes of data, containing more than 20K malware samples. The analysis
of this dataset inspired the development of a novel paradigm that is effective
in categorizing malware variants into their actual family groups. This paradigm
is presented and discussed in the present paper, where emphasis has been given
to the phases related to the extraction, and selection of a set of novel
features for the effective representation of malware samples. Features can be
grouped according to different characteristics of malware behavior, and their
fusion is performed according to a per-class weighting paradigm. The proposed
method achieved a very high accuracy ($\approx$ 0.998) on the Microsoft Malware
Challenge dataset.
| [
{
"version": "v1",
"created": "Fri, 13 Nov 2015 15:33:02 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Mar 2016 10:21:15 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Ahmadi",
"Mansour",
""
],
[
"Ulyanov",
"Dmitry",
""
],
[
"Semenov",
"Stanislav",
""
],
[
"Trofimov",
"Mikhail",
""
],
[
"Giacinto",
"Giorgio",
""
]
] | TITLE: Novel Feature Extraction, Selection and Fusion for Effective Malware
Family Classification
ABSTRACT: Modern malware is designed with mutation characteristics, namely polymorphism
and metamorphism, which causes an enormous growth in the number of variants of
malware samples. Categorization of malware samples on the basis of their
behaviors is essential for the computer security community, because they
receive huge number of malware everyday, and the signature extraction process
is usually based on malicious parts characterizing malware families. Microsoft
released a malware classification challenge in 2015 with a huge dataset of near
0.5 terabytes of data, containing more than 20K malware samples. The analysis
of this dataset inspired the development of a novel paradigm that is effective
in categorizing malware variants into their actual family groups. This paradigm
is presented and discussed in the present paper, where emphasis has been given
to the phases related to the extraction, and selection of a set of novel
features for the effective representation of malware samples. Features can be
grouped according to different characteristics of malware behavior, and their
fusion is performed according to a per-class weighting paradigm. The proposed
method achieved a very high accuracy ($\approx$ 0.998) on the Microsoft Malware
Challenge dataset.
| new_dataset | 0.845815 |
1512.07372 | Pin-Yu Chen | Pin-Yu Chen, Sutanay Choudhury, Alfred O. Hero | Multi-centrality Graph Spectral Decompositions and their Application to
Cyber Intrusion Detection | To appear in ICASSP 2016 | null | null | null | cs.SI cs.CR stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many modern datasets can be represented as graphs and hence spectral
decompositions such as graph principal component analysis (PCA) can be useful.
Distinct from previous graph decomposition approaches based on subspace
projection of a single topological feature, e.g., the Fiedler vector of
centered graph adjacency matrix (graph Laplacian), we propose spectral
decomposition approaches to graph PCA and graph dictionary learning that
integrate multiple features, including graph walk statistics, centrality
measures and graph distances to reference nodes. In this paper we propose a new
PCA method for single graph analysis, called multi-centrality graph PCA
(MC-GPCA), and a new dictionary learning method for ensembles of graphs, called
multi-centrality graph dictionary learning (MC-GDL), both based on spectral
decomposition of multi-centrality matrices. As an application to cyber
intrusion detection, MC-GPCA can be an effective indicator of anomalous
connectivity pattern and MC-GDL can provide discriminative basis for attack
classification.
| [
{
"version": "v1",
"created": "Wed, 23 Dec 2015 07:13:27 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Mar 2016 03:31:25 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Chen",
"Pin-Yu",
""
],
[
"Choudhury",
"Sutanay",
""
],
[
"Hero",
"Alfred O.",
""
]
] | TITLE: Multi-centrality Graph Spectral Decompositions and their Application to
Cyber Intrusion Detection
ABSTRACT: Many modern datasets can be represented as graphs and hence spectral
decompositions such as graph principal component analysis (PCA) can be useful.
Distinct from previous graph decomposition approaches based on subspace
projection of a single topological feature, e.g., the Fiedler vector of
centered graph adjacency matrix (graph Laplacian), we propose spectral
decomposition approaches to graph PCA and graph dictionary learning that
integrate multiple features, including graph walk statistics, centrality
measures and graph distances to reference nodes. In this paper we propose a new
PCA method for single graph analysis, called multi-centrality graph PCA
(MC-GPCA), and a new dictionary learning method for ensembles of graphs, called
multi-centrality graph dictionary learning (MC-GDL), both based on spectral
decomposition of multi-centrality matrices. As an application to cyber
intrusion detection, MC-GPCA can be an effective indicator of anomalous
connectivity pattern and MC-GDL can provide discriminative basis for attack
classification.
| no_new_dataset | 0.94474 |
1603.03097 | Yu Wang | Yu Wang, Yuncheng Li, Jiebo Luo | Deciphering the 2016 U.S. Presidential Campaign in the Twitter Sphere: A
Comparison of the Trumpists and Clintonists | 4 pages, to appear in the 10th International AAAI Conference on Web
and Social Media | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study follower demographics of Donald Trump and Hillary
Clinton, the two leading candidates in the 2016 U.S. presidential race. We
build a unique dataset US2016, which includes the number of followers for each
candidate from September 17, 2015 to December 22, 2015. US2016 also includes
the geographical location of these followers, the number of their own followers
and, very importantly, the profile image of each follower. We use individuals'
number of followers and profile images to analyze four dimensions of follower
demographics: social status, gender, race and age. Our study shows that in
terms of social influence, the Trumpists are more polarized than the
Clintonists: they tend to have either a lot of influence or little influence.
We also find that compared with the Clintonists, the Trumpists are more likely
to be either very young or very old. Our study finds no gender affinity effect
for Clinton in the Twitter sphere, but we do find that the Clintonists are more
racially diverse.
| [
{
"version": "v1",
"created": "Wed, 9 Mar 2016 23:36:26 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Wang",
"Yu",
""
],
[
"Li",
"Yuncheng",
""
],
[
"Luo",
"Jiebo",
""
]
] | TITLE: Deciphering the 2016 U.S. Presidential Campaign in the Twitter Sphere: A
Comparison of the Trumpists and Clintonists
ABSTRACT: In this paper, we study follower demographics of Donald Trump and Hillary
Clinton, the two leading candidates in the 2016 U.S. presidential race. We
build a unique dataset US2016, which includes the number of followers for each
candidate from September 17, 2015 to December 22, 2015. US2016 also includes
the geographical location of these followers, the number of their own followers
and, very importantly, the profile image of each follower. We use individuals'
number of followers and profile images to analyze four dimensions of follower
demographics: social status, gender, race and age. Our study shows that in
terms of social influence, the Trumpists are more polarized than the
Clintonists: they tend to have either a lot of influence or little influence.
We also find that compared with the Clintonists, the Trumpists are more likely
to be either very young or very old. Our study finds no gender affinity effect
for Clinton in the Twitter sphere, but we do find that the Clintonists are more
racially diverse.
| new_dataset | 0.956796 |
1603.03101 | Chen-Yu Lee | Chen-Yu Lee and Simon Osindero | Recursive Recurrent Nets with Attention Modeling for OCR in the Wild | accepted at CVPR 2016 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present recursive recurrent neural networks with attention modeling
(R$^2$AM) for lexicon-free optical character recognition in natural scene
images. The primary advantages of the proposed method are: (1) use of recursive
convolutional neural networks (CNNs), which allow for parametrically efficient
and effective image feature extraction; (2) an implicitly learned
character-level language model, embodied in a recurrent neural network which
avoids the need to use N-grams; and (3) the use of a soft-attention mechanism,
allowing the model to selectively exploit image features in a coordinated way,
and allowing for end-to-end training within a standard backpropagation
framework. We validate our method with state-of-the-art performance on
challenging benchmark datasets: Street View Text, IIIT5k, ICDAR and Synth90k.
| [
{
"version": "v1",
"created": "Wed, 9 Mar 2016 23:49:51 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Lee",
"Chen-Yu",
""
],
[
"Osindero",
"Simon",
""
]
] | TITLE: Recursive Recurrent Nets with Attention Modeling for OCR in the Wild
ABSTRACT: We present recursive recurrent neural networks with attention modeling
(R$^2$AM) for lexicon-free optical character recognition in natural scene
images. The primary advantages of the proposed method are: (1) use of recursive
convolutional neural networks (CNNs), which allow for parametrically efficient
and effective image feature extraction; (2) an implicitly learned
character-level language model, embodied in a recurrent neural network which
avoids the need to use N-grams; and (3) the use of a soft-attention mechanism,
allowing the model to selectively exploit image features in a coordinated way,
and allowing for end-to-end training within a standard backpropagation
framework. We validate our method with state-of-the-art performance on
challenging benchmark datasets: Street View Text, IIIT5k, ICDAR and Synth90k.
| no_new_dataset | 0.953923 |
1603.03281 | UshaRani Yelipe | Yelipe UshaRani, P. Sammulal | An Innovative Imputation and Classification Approach for Accurate
Disease Prediction | Special Issue of Journal IJCSIS indexed in Web of Science and Thomson
Reuters ISI. https://sites.google.com/site/ijcsis/vol-14-s1-feb-2016 | null | null | null | cs.DB cs.IR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Imputation of missing attribute values in medical datasets for extracting
hidden knowledge from medical datasets is an interesting research topic of
interest which is very challenging. One cannot eliminate missing values in
medical records. The reason may be because some tests may not been conducted as
they are cost effective, values missed when conducting clinical trials, values
may not have been recorded to name some of the reasons. Data mining researchers
have been proposing various approaches to find and impute missing values to
increase classification accuracies so that disease may be predicted accurately.
In this paper, we propose a novel imputation approach for imputation of missing
values and performing classification after fixing missing values. The approach
is based on clustering concept and aims at dimensionality reduction of the
records. The case study discussed shows that missing values can be fixed and
imputed efficiently by achieving dimensionality reduction. The importance of
proposed approach for classification is visible in the case study which assigns
single class label in contrary to multi-label assignment if dimensionality
reduction is not performed.
| [
{
"version": "v1",
"created": "Thu, 10 Mar 2016 14:31:33 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"UshaRani",
"Yelipe",
""
],
[
"Sammulal",
"P.",
""
]
] | TITLE: An Innovative Imputation and Classification Approach for Accurate
Disease Prediction
ABSTRACT: Imputation of missing attribute values in medical datasets for extracting
hidden knowledge from medical datasets is an interesting research topic of
interest which is very challenging. One cannot eliminate missing values in
medical records. The reason may be because some tests may not been conducted as
they are cost effective, values missed when conducting clinical trials, values
may not have been recorded to name some of the reasons. Data mining researchers
have been proposing various approaches to find and impute missing values to
increase classification accuracies so that disease may be predicted accurately.
In this paper, we propose a novel imputation approach for imputation of missing
values and performing classification after fixing missing values. The approach
is based on clustering concept and aims at dimensionality reduction of the
records. The case study discussed shows that missing values can be fixed and
imputed efficiently by achieving dimensionality reduction. The importance of
proposed approach for classification is visible in the case study which assigns
single class label in contrary to multi-label assignment if dimensionality
reduction is not performed.
| no_new_dataset | 0.947575 |
1603.03303 | Rahmtin Rotabi | Rahmtin Rotabi and Jon Kleinberg | The Status Gradient of Trends in Social Media | null | null | null | null | cs.SI cs.CY physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An active line of research has studied the detection and representation of
trends in social media content. There is still relatively little understanding,
however, of methods to characterize the early adopters of these trends: who
picks up on these trends at different points in time, and what is their role in
the system? We develop a framework for analyzing the population of users who
participate in trending topics over the course of these topics' lifecycles.
Central to our analysis is the notion of a "status gradient", describing how
users of different activity levels adopt a trend at different points in time.
Across multiple datasets, we find that this methodology reveals key differences
in the nature of the early adopters in different domains.
| [
{
"version": "v1",
"created": "Thu, 10 Mar 2016 15:46:07 GMT"
}
] | 2016-03-11T00:00:00 | [
[
"Rotabi",
"Rahmtin",
""
],
[
"Kleinberg",
"Jon",
""
]
] | TITLE: The Status Gradient of Trends in Social Media
ABSTRACT: An active line of research has studied the detection and representation of
trends in social media content. There is still relatively little understanding,
however, of methods to characterize the early adopters of these trends: who
picks up on these trends at different points in time, and what is their role in
the system? We develop a framework for analyzing the population of users who
participate in trending topics over the course of these topics' lifecycles.
Central to our analysis is the notion of a "status gradient", describing how
users of different activity levels adopt a trend at different points in time.
Across multiple datasets, we find that this methodology reveals key differences
in the nature of the early adopters in different domains.
| no_new_dataset | 0.943971 |
1504.07947 | Le Hou | Le Hou, Dimitris Samaras, Tahsin M. Kurc, Yi Gao, James E. Davis, Joel
H. Saltz | Patch-based Convolutional Neural Network for Whole Slide Tissue Image
Classification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Convolutional Neural Networks (CNN) are state-of-the-art models for many
image classification tasks. However, to recognize cancer subtypes
automatically, training a CNN on gigapixel resolution Whole Slide Tissue Images
(WSI) is currently computationally impossible. The differentiation of cancer
subtypes is based on cellular-level visual features observed on image patch
scale. Therefore, we argue that in this situation, training a patch-level
classifier on image patches will perform better than or similar to an
image-level classifier. The challenge becomes how to intelligently combine
patch-level classification results and model the fact that not all patches will
be discriminative. We propose to train a decision fusion model to aggregate
patch-level predictions given by patch-level CNNs, which to the best of our
knowledge has not been shown before. Furthermore, we formulate a novel
Expectation-Maximization (EM) based method that automatically locates
discriminative patches robustly by utilizing the spatial relationships of
patches. We apply our method to the classification of glioma and non-small-cell
lung carcinoma cases into subtypes. The classification accuracy of our method
is similar to the inter-observer agreement between pathologists. Although it is
impossible to train CNNs on WSIs, we experimentally demonstrate using a
comparable non-cancer dataset of smaller images that a patch-based CNN can
outperform an image-based CNN.
| [
{
"version": "v1",
"created": "Wed, 29 Apr 2015 18:15:22 GMT"
},
{
"version": "v2",
"created": "Mon, 11 May 2015 01:55:55 GMT"
},
{
"version": "v3",
"created": "Tue, 19 May 2015 21:01:11 GMT"
},
{
"version": "v4",
"created": "Tue, 8 Mar 2016 18:07:03 GMT"
},
{
"version": "v5",
"created": "Wed, 9 Mar 2016 14:26:16 GMT"
}
] | 2016-03-10T00:00:00 | [
[
"Hou",
"Le",
""
],
[
"Samaras",
"Dimitris",
""
],
[
"Kurc",
"Tahsin M.",
""
],
[
"Gao",
"Yi",
""
],
[
"Davis",
"James E.",
""
],
[
"Saltz",
"Joel H.",
""
]
] | TITLE: Patch-based Convolutional Neural Network for Whole Slide Tissue Image
Classification
ABSTRACT: Convolutional Neural Networks (CNN) are state-of-the-art models for many
image classification tasks. However, to recognize cancer subtypes
automatically, training a CNN on gigapixel resolution Whole Slide Tissue Images
(WSI) is currently computationally impossible. The differentiation of cancer
subtypes is based on cellular-level visual features observed on image patch
scale. Therefore, we argue that in this situation, training a patch-level
classifier on image patches will perform better than or similar to an
image-level classifier. The challenge becomes how to intelligently combine
patch-level classification results and model the fact that not all patches will
be discriminative. We propose to train a decision fusion model to aggregate
patch-level predictions given by patch-level CNNs, which to the best of our
knowledge has not been shown before. Furthermore, we formulate a novel
Expectation-Maximization (EM) based method that automatically locates
discriminative patches robustly by utilizing the spatial relationships of
patches. We apply our method to the classification of glioma and non-small-cell
lung carcinoma cases into subtypes. The classification accuracy of our method
is similar to the inter-observer agreement between pathologists. Although it is
impossible to train CNNs on WSIs, we experimentally demonstrate using a
comparable non-cancer dataset of smaller images that a patch-based CNN can
outperform an image-based CNN.
| no_new_dataset | 0.951818 |
1506.01273 | David Martins de Matos | Marta Apar\'icio, Paulo Figueiredo, Francisco Raposo, David Martins de
Matos, Ricardo Ribeiro, Lu\'is Marujo | Summarization of Films and Documentaries Based on Subtitles and Scripts | 7 pages, 9 tables, 4 figures, submitted to Pattern Recognition
Letters (Elsevier) | Pattern Recognition Letters, Volume 73, 1 April 2016, Pages 7-12 | 10.1016/j.patrec.2015.12.016 | null | cs.CL cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We assess the performance of generic text summarization algorithms applied to
films and documentaries, using the well-known behavior of summarization of news
articles as reference. We use three datasets: (i) news articles, (ii) film
scripts and subtitles, and (iii) documentary subtitles. Standard ROUGE metrics
are used for comparing generated summaries against news abstracts, plot
summaries, and synopses. We show that the best performing algorithms are LSA,
for news articles and documentaries, and LexRank and Support Sets, for films.
Despite the different nature of films and documentaries, their relative
behavior is in accordance with that obtained for news articles.
| [
{
"version": "v1",
"created": "Wed, 3 Jun 2015 15:07:14 GMT"
},
{
"version": "v2",
"created": "Thu, 4 Jun 2015 12:41:55 GMT"
},
{
"version": "v3",
"created": "Wed, 9 Mar 2016 16:50:43 GMT"
}
] | 2016-03-10T00:00:00 | [
[
"Aparício",
"Marta",
""
],
[
"Figueiredo",
"Paulo",
""
],
[
"Raposo",
"Francisco",
""
],
[
"de Matos",
"David Martins",
""
],
[
"Ribeiro",
"Ricardo",
""
],
[
"Marujo",
"Luís",
""
]
] | TITLE: Summarization of Films and Documentaries Based on Subtitles and Scripts
ABSTRACT: We assess the performance of generic text summarization algorithms applied to
films and documentaries, using the well-known behavior of summarization of news
articles as reference. We use three datasets: (i) news articles, (ii) film
scripts and subtitles, and (iii) documentary subtitles. Standard ROUGE metrics
are used for comparing generated summaries against news abstracts, plot
summaries, and synopses. We show that the best performing algorithms are LSA,
for news articles and documentaries, and LexRank and Support Sets, for films.
Despite the different nature of films and documentaries, their relative
behavior is in accordance with that obtained for news articles.
| no_new_dataset | 0.947962 |
1602.00991 | Peter Ondruska | Peter Ondruska and Ingmar Posner | Deep Tracking: Seeing Beyond Seeing Using Recurrent Neural Networks | Published in The Thirtieth AAAI Conference on Artificial Intelligence
(AAAI-16), Video: https://youtu.be/cdeWCpfUGWc, Code:
http://mrg.robots.ox.ac.uk/mrg_people/peter-ondruska/ | null | null | null | cs.LG cs.AI cs.CV cs.NE cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents to the best of our knowledge the first end-to-end object
tracking approach which directly maps from raw sensor input to object tracks in
sensor space without requiring any feature engineering or system identification
in the form of plant or sensor models. Specifically, our system accepts a
stream of raw sensor data at one end and, in real-time, produces an estimate of
the entire environment state at the output including even occluded objects. We
achieve this by framing the problem as a deep learning task and exploit
sequence models in the form of recurrent neural networks to learn a mapping
from sensor measurements to object tracks. In particular, we propose a learning
method based on a form of input dropout which allows learning in an
unsupervised manner, only based on raw, occluded sensor data without access to
ground-truth annotations. We demonstrate our approach using a synthetic dataset
designed to mimic the task of tracking objects in 2D laser data -- as commonly
encountered in robotics applications -- and show that it learns to track many
dynamic objects despite occlusions and the presence of sensor noise.
| [
{
"version": "v1",
"created": "Tue, 2 Feb 2016 16:10:16 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Mar 2016 22:09:05 GMT"
}
] | 2016-03-10T00:00:00 | [
[
"Ondruska",
"Peter",
""
],
[
"Posner",
"Ingmar",
""
]
] | TITLE: Deep Tracking: Seeing Beyond Seeing Using Recurrent Neural Networks
ABSTRACT: This paper presents to the best of our knowledge the first end-to-end object
tracking approach which directly maps from raw sensor input to object tracks in
sensor space without requiring any feature engineering or system identification
in the form of plant or sensor models. Specifically, our system accepts a
stream of raw sensor data at one end and, in real-time, produces an estimate of
the entire environment state at the output including even occluded objects. We
achieve this by framing the problem as a deep learning task and exploit
sequence models in the form of recurrent neural networks to learn a mapping
from sensor measurements to object tracks. In particular, we propose a learning
method based on a form of input dropout which allows learning in an
unsupervised manner, only based on raw, occluded sensor data without access to
ground-truth annotations. We demonstrate our approach using a synthetic dataset
designed to mimic the task of tracking objects in 2D laser data -- as commonly
encountered in robotics applications -- and show that it learns to track many
dynamic objects despite occlusions and the presence of sensor noise.
| new_dataset | 0.965381 |
1603.02869 | Emanuele Lindo Secco | Daniel Elstob, Emanuele Lindo Secco | A Low Cost Eeg Based Bci Prosthetic Using Motor Imagery | International Journal of Information Technology Convergence and
Services (IJITCS) Vol.6, No.1,February 2016 | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Brain Computer Interfaces (BCI) provide the opportunity to control external
devices using the brain ElectroEncephaloGram (EEG) signals. In this paper we
propose two software framework in order to control a 5 degree of freedom
robotic and prosthetic hand. Results are presented where an Emotiv Cognitive
Suite (i.e. the 1st framework) combined with an embedded software system (i.e.
an open source Arduino board) is able to control the hand through character
input associated with the taught actions of the suite. This system provides
evidence of the feasibility of brain signals being a viable approach to
controlling the chosen prosthetic. Results are then presented in the second
framework. This latter one allowed for the training and classification of EEG
signals for motor imagery tasks. When analysing the system, clear visual
representations of the performance and accuracy are presented in the results
using a confusion matrix, accuracy measurement and a feedback bar signifying
signal strength. Experiments with various acquisition datasets were carried out
and with a critical evaluation of the results given. Finally depending on the
classification of the brain signal a Python script outputs the driving command
to the Arduino to control the prosthetic. The proposed architecture performs
overall good results for the design and implementation of economically
convenient BCI and prosthesis.
| [
{
"version": "v1",
"created": "Wed, 9 Mar 2016 12:39:06 GMT"
}
] | 2016-03-10T00:00:00 | [
[
"Elstob",
"Daniel",
""
],
[
"Secco",
"Emanuele Lindo",
""
]
] | TITLE: A Low Cost Eeg Based Bci Prosthetic Using Motor Imagery
ABSTRACT: Brain Computer Interfaces (BCI) provide the opportunity to control external
devices using the brain ElectroEncephaloGram (EEG) signals. In this paper we
propose two software framework in order to control a 5 degree of freedom
robotic and prosthetic hand. Results are presented where an Emotiv Cognitive
Suite (i.e. the 1st framework) combined with an embedded software system (i.e.
an open source Arduino board) is able to control the hand through character
input associated with the taught actions of the suite. This system provides
evidence of the feasibility of brain signals being a viable approach to
controlling the chosen prosthetic. Results are then presented in the second
framework. This latter one allowed for the training and classification of EEG
signals for motor imagery tasks. When analysing the system, clear visual
representations of the performance and accuracy are presented in the results
using a confusion matrix, accuracy measurement and a feedback bar signifying
signal strength. Experiments with various acquisition datasets were carried out
and with a critical evaluation of the results given. Finally depending on the
classification of the brain signal a Python script outputs the driving command
to the Arduino to control the prosthetic. The proposed architecture performs
overall good results for the design and implementation of economically
convenient BCI and prosthesis.
| no_new_dataset | 0.949248 |
1312.6947 | Sourish Dasgupta | Sourish Dasgupta, Ankur Padia, Kushal Shah, Prasenjit Majumder | Formal Ontology Learning on Factual IS-A Corpus in English using
Description Logics | This paper has been withdrawn by the author due to requirement of
re-evaluation of results | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ontology Learning (OL) is the computational task of generating a knowledge
base in the form of an ontology given an unstructured corpus whose content is
in natural language (NL). Several works can be found in this area most of which
are limited to statistical and lexico-syntactic pattern matching based
techniques Light-Weight OL. These techniques do not lead to very accurate
learning mostly because of several linguistic nuances in NL. Formal OL is an
alternative (less explored) methodology were deep linguistics analysis is made
using theory and tools found in computational linguistics to generate formal
axioms and definitions instead simply inducing a taxonomy. In this paper we
propose "Description Logic (DL)" based formal OL framework for learning factual
IS-A type sentences in English. We claim that semantic construction of IS-A
sentences is non trivial. Hence, we also claim that such sentences requires
special studies in the context of OL before any truly formal OL can be
proposed. We introduce a learner tool, called DLOL_IS-A, that generated such
ontologies in the owl format. We have adopted "Gold Standard" based OL
evaluation on IS-A rich WCL v.1.1 dataset and our own Community representative
IS-A dataset. We observed significant improvement of DLOL_IS-A when compared to
the light-weight OL tool Text2Onto and formal OL tool FRED.
| [
{
"version": "v1",
"created": "Wed, 25 Dec 2013 09:17:28 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Mar 2016 05:20:47 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Dasgupta",
"Sourish",
""
],
[
"Padia",
"Ankur",
""
],
[
"Shah",
"Kushal",
""
],
[
"Majumder",
"Prasenjit",
""
]
] | TITLE: Formal Ontology Learning on Factual IS-A Corpus in English using
Description Logics
ABSTRACT: Ontology Learning (OL) is the computational task of generating a knowledge
base in the form of an ontology given an unstructured corpus whose content is
in natural language (NL). Several works can be found in this area most of which
are limited to statistical and lexico-syntactic pattern matching based
techniques Light-Weight OL. These techniques do not lead to very accurate
learning mostly because of several linguistic nuances in NL. Formal OL is an
alternative (less explored) methodology were deep linguistics analysis is made
using theory and tools found in computational linguistics to generate formal
axioms and definitions instead simply inducing a taxonomy. In this paper we
propose "Description Logic (DL)" based formal OL framework for learning factual
IS-A type sentences in English. We claim that semantic construction of IS-A
sentences is non trivial. Hence, we also claim that such sentences requires
special studies in the context of OL before any truly formal OL can be
proposed. We introduce a learner tool, called DLOL_IS-A, that generated such
ontologies in the owl format. We have adopted "Gold Standard" based OL
evaluation on IS-A rich WCL v.1.1 dataset and our own Community representative
IS-A dataset. We observed significant improvement of DLOL_IS-A when compared to
the light-weight OL tool Text2Onto and formal OL tool FRED.
| new_dataset | 0.878783 |
1601.04814 | Gianmarco De Francisci Morales | Gianmarco De Francisci Morales and Aristides Gionis | Streaming Similarity Self-Join | null | null | null | null | cs.DB cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce and study the problem of computing the similarity self-join in a
streaming context (SSSJ), where the input is an unbounded stream of items
arriving continuously. The goal is to find all pairs of items in the stream
whose similarity is greater than a given threshold. The simplest formulation of
the problem requires unbounded memory, and thus, it is intractable. To make the
problem feasible, we introduce the notion of time-dependent similarity: the
similarity of two items decreases with the difference in their arrival time. By
leveraging the properties of this time-dependent similarity function, we design
two algorithmic frameworks to solve the sssj problem. The first one, MiniBatch
(MB), uses existing index-based filtering techniques for the static version of
the problem, and combines them in a pipeline. The second framework, Streaming
(STR), adds time filtering to the existing indexes, and integrates new
time-based bounds deeply in the working of the algorithms. We also introduce a
new indexing technique (L2), which is based on an existing state-of-the-art
indexing technique (L2AP), but is optimized for the streaming case. Extensive
experiments show that the STR algorithm, when instantiated with the L2 index,
is the most scalable option across a wide array of datasets and parameters.
| [
{
"version": "v1",
"created": "Tue, 19 Jan 2016 07:34:17 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Mar 2016 09:14:27 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Morales",
"Gianmarco De Francisci",
""
],
[
"Gionis",
"Aristides",
""
]
] | TITLE: Streaming Similarity Self-Join
ABSTRACT: We introduce and study the problem of computing the similarity self-join in a
streaming context (SSSJ), where the input is an unbounded stream of items
arriving continuously. The goal is to find all pairs of items in the stream
whose similarity is greater than a given threshold. The simplest formulation of
the problem requires unbounded memory, and thus, it is intractable. To make the
problem feasible, we introduce the notion of time-dependent similarity: the
similarity of two items decreases with the difference in their arrival time. By
leveraging the properties of this time-dependent similarity function, we design
two algorithmic frameworks to solve the sssj problem. The first one, MiniBatch
(MB), uses existing index-based filtering techniques for the static version of
the problem, and combines them in a pipeline. The second framework, Streaming
(STR), adds time filtering to the existing indexes, and integrates new
time-based bounds deeply in the working of the algorithms. We also introduce a
new indexing technique (L2), which is based on an existing state-of-the-art
indexing technique (L2AP), but is optimized for the streaming case. Extensive
experiments show that the STR algorithm, when instantiated with the L2 index,
is the most scalable option across a wide array of datasets and parameters.
| no_new_dataset | 0.946745 |
1601.06892 | Kuldeep S Kulkarni Mr. | Kuldeep Kulkarni, Suhas Lohit, Pavan Turaga, Ronan Kerviche, Amit
Ashok | ReconNet: Non-Iterative Reconstruction of Images from Compressively
Sensed Random Measurements | Accepted at IEEE International Conference on Computer Vision and
Pattern Recognition (CVPR), 2016 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The goal of this paper is to present a non-iterative and more importantly an
extremely fast algorithm to reconstruct images from compressively sensed (CS)
random measurements. To this end, we propose a novel convolutional neural
network (CNN) architecture which takes in CS measurements of an image as input
and outputs an intermediate reconstruction. We call this network, ReconNet. The
intermediate reconstruction is fed into an off-the-shelf denoiser to obtain the
final reconstructed image. On a standard dataset of images we show significant
improvements in reconstruction results (both in terms of PSNR and time
complexity) over state-of-the-art iterative CS reconstruction algorithms at
various measurement rates. Further, through qualitative experiments on real
data collected using our block single pixel camera (SPC), we show that our
network is highly robust to sensor noise and can recover visually better
quality images than competitive algorithms at extremely low sensing rates of
0.1 and 0.04. To demonstrate that our algorithm can recover semantically
informative images even at a low measurement rate of 0.01, we present a very
robust proof of concept real-time visual tracking application.
| [
{
"version": "v1",
"created": "Tue, 26 Jan 2016 05:17:14 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Mar 2016 23:31:08 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Kulkarni",
"Kuldeep",
""
],
[
"Lohit",
"Suhas",
""
],
[
"Turaga",
"Pavan",
""
],
[
"Kerviche",
"Ronan",
""
],
[
"Ashok",
"Amit",
""
]
] | TITLE: ReconNet: Non-Iterative Reconstruction of Images from Compressively
Sensed Random Measurements
ABSTRACT: The goal of this paper is to present a non-iterative and more importantly an
extremely fast algorithm to reconstruct images from compressively sensed (CS)
random measurements. To this end, we propose a novel convolutional neural
network (CNN) architecture which takes in CS measurements of an image as input
and outputs an intermediate reconstruction. We call this network, ReconNet. The
intermediate reconstruction is fed into an off-the-shelf denoiser to obtain the
final reconstructed image. On a standard dataset of images we show significant
improvements in reconstruction results (both in terms of PSNR and time
complexity) over state-of-the-art iterative CS reconstruction algorithms at
various measurement rates. Further, through qualitative experiments on real
data collected using our block single pixel camera (SPC), we show that our
network is highly robust to sensor noise and can recover visually better
quality images than competitive algorithms at extremely low sensing rates of
0.1 and 0.04. To demonstrate that our algorithm can recover semantically
informative images even at a low measurement rate of 0.01, we present a very
robust proof of concept real-time visual tracking application.
| no_new_dataset | 0.950227 |
1602.06866 | Hao Wu | Huijuan Shao, K.S.M. Tozammel Hossain, Hao Wu, Maleq Khan, Anil
Vullikanti, B. Aditya Prakash, Madhav Marathe and Naren Ramakrishnan | Forecasting the Flu: Designing Social Network Sensors for Epidemics | The conference version of the paper is submitted for publication | null | null | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Early detection and modeling of a contagious epidemic can provide important
guidance about quelling the contagion, controlling its spread, or the effective
design of countermeasures. A topic of recent interest has been to design social
network sensors, i.e., identifying a small set of people who can be monitored
to provide insight into the emergence of an epidemic in a larger population. We
formally pose the problem of designing social network sensors for flu epidemics
and identify two different objectives that could be targeted in such sensor
design problems. Using the graph theoretic notion of dominators we develop an
efficient and effective heuristic for forecasting epidemics at lead time. Using
six city-scale datasets generated by extensive microscopic epidemiological
simulations involving millions of individuals, we illustrate the practical
applicability of our methods and show significant benefits (up to twenty-two
days more lead time) compared to other competitors. Most importantly, we
demonstrate the use of surrogates or proxies for policy makers for designing
social network sensors that require from nonintrusive knowledge of people to
more information on the relationship among people. The results show that the
more intrusive information we obtain, the longer lead time to predict the flu
outbreak up to nine days.
| [
{
"version": "v1",
"created": "Mon, 22 Feb 2016 17:32:31 GMT"
},
{
"version": "v2",
"created": "Thu, 25 Feb 2016 17:20:49 GMT"
},
{
"version": "v3",
"created": "Tue, 8 Mar 2016 16:46:37 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Shao",
"Huijuan",
""
],
[
"Hossain",
"K. S. M. Tozammel",
""
],
[
"Wu",
"Hao",
""
],
[
"Khan",
"Maleq",
""
],
[
"Vullikanti",
"Anil",
""
],
[
"Prakash",
"B. Aditya",
""
],
[
"Marathe",
"Madhav",
""
],
[
"Ramakrishnan",
"Naren",
""
]
] | TITLE: Forecasting the Flu: Designing Social Network Sensors for Epidemics
ABSTRACT: Early detection and modeling of a contagious epidemic can provide important
guidance about quelling the contagion, controlling its spread, or the effective
design of countermeasures. A topic of recent interest has been to design social
network sensors, i.e., identifying a small set of people who can be monitored
to provide insight into the emergence of an epidemic in a larger population. We
formally pose the problem of designing social network sensors for flu epidemics
and identify two different objectives that could be targeted in such sensor
design problems. Using the graph theoretic notion of dominators we develop an
efficient and effective heuristic for forecasting epidemics at lead time. Using
six city-scale datasets generated by extensive microscopic epidemiological
simulations involving millions of individuals, we illustrate the practical
applicability of our methods and show significant benefits (up to twenty-two
days more lead time) compared to other competitors. Most importantly, we
demonstrate the use of surrogates or proxies for policy makers for designing
social network sensors that require from nonintrusive knowledge of people to
more information on the relationship among people. The results show that the
more intrusive information we obtain, the longer lead time to predict the flu
outbreak up to nine days.
| no_new_dataset | 0.9463 |
1603.00893 | Boxiang Dong | Boxiang Dong, Hui Wendy Wang | Frequency-hiding Dependency-preserving Encryption for Outsourced
Databases | null | null | null | null | cs.DB cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The cloud paradigm enables users to outsource their data to computationally
powerful third-party service providers for data management. Many data
management tasks rely on the data dependencies in the outsourced data. This
raises an important issue of how the data owner can protect the sensitive
information in the outsourced data while preserving the data dependencies. In
this paper, we consider functional dependency FD, an important type of data
dependency. We design a FD-preserving encryption scheme, named F2, that enables
the service provider to discover the FDs from the encrypted dataset. We
consider the frequency analysis attack, and show that the F2 encryption scheme
can defend against the attack under Kerckhoff's principle with provable
guarantee. Our empirical study demonstrates the efficiency and effectiveness of
F2.
| [
{
"version": "v1",
"created": "Wed, 2 Mar 2016 21:20:16 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Mar 2016 15:16:00 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Dong",
"Boxiang",
""
],
[
"Wang",
"Hui Wendy",
""
]
] | TITLE: Frequency-hiding Dependency-preserving Encryption for Outsourced
Databases
ABSTRACT: The cloud paradigm enables users to outsource their data to computationally
powerful third-party service providers for data management. Many data
management tasks rely on the data dependencies in the outsourced data. This
raises an important issue of how the data owner can protect the sensitive
information in the outsourced data while preserving the data dependencies. In
this paper, we consider functional dependency FD, an important type of data
dependency. We design a FD-preserving encryption scheme, named F2, that enables
the service provider to discover the FDs from the encrypted dataset. We
consider the frequency analysis attack, and show that the F2 encryption scheme
can defend against the attack under Kerckhoff's principle with provable
guarantee. Our empirical study demonstrates the efficiency and effectiveness of
F2.
| no_new_dataset | 0.949576 |
1603.01670 | Tao Wei | Tao Wei, Changhu Wang, Yong Rui, Chang Wen Chen | Network Morphism | Under review for ICML 2016 | null | null | null | cs.LG cs.CV cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present in this paper a systematic study on how to morph a well-trained
neural network to a new one so that its network function can be completely
preserved. We define this as \emph{network morphism} in this research. After
morphing a parent network, the child network is expected to inherit the
knowledge from its parent network and also has the potential to continue
growing into a more powerful one with much shortened training time. The first
requirement for this network morphism is its ability to handle diverse morphing
types of networks, including changes of depth, width, kernel size, and even
subnet. To meet this requirement, we first introduce the network morphism
equations, and then develop novel morphing algorithms for all these morphing
types for both classic and convolutional neural networks. The second
requirement for this network morphism is its ability to deal with non-linearity
in a network. We propose a family of parametric-activation functions to
facilitate the morphing of any continuous non-linear activation neurons.
Experimental results on benchmark datasets and typical neural networks
demonstrate the effectiveness of the proposed network morphism scheme.
| [
{
"version": "v1",
"created": "Sat, 5 Mar 2016 02:06:43 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Mar 2016 16:36:00 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Wei",
"Tao",
""
],
[
"Wang",
"Changhu",
""
],
[
"Rui",
"Yong",
""
],
[
"Chen",
"Chang Wen",
""
]
] | TITLE: Network Morphism
ABSTRACT: We present in this paper a systematic study on how to morph a well-trained
neural network to a new one so that its network function can be completely
preserved. We define this as \emph{network morphism} in this research. After
morphing a parent network, the child network is expected to inherit the
knowledge from its parent network and also has the potential to continue
growing into a more powerful one with much shortened training time. The first
requirement for this network morphism is its ability to handle diverse morphing
types of networks, including changes of depth, width, kernel size, and even
subnet. To meet this requirement, we first introduce the network morphism
equations, and then develop novel morphing algorithms for all these morphing
types for both classic and convolutional neural networks. The second
requirement for this network morphism is its ability to deal with non-linearity
in a network. We propose a family of parametric-activation functions to
facilitate the morphing of any continuous non-linear activation neurons.
Experimental results on benchmark datasets and typical neural networks
demonstrate the effectiveness of the proposed network morphism scheme.
| no_new_dataset | 0.953923 |
1603.02494 | Tapesh Santra | Tapesh Santra | A Bayesian non-parametric method for clustering high-dimensional binary
data | null | null | null | null | stat.AP cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In many real life problems, objects are described by large number of binary
features. For instance, documents are characterized by presence or absence of
certain keywords; cancer patients are characterized by presence or absence of
certain mutations etc. In such cases, grouping together similar
objects/profiles based on such high dimensional binary features is desirable,
but challenging. Here, I present a Bayesian non parametric algorithm for
clustering high dimensional binary data. It uses a Dirichlet Process (DP)
mixture model and simulated annealing to not only cluster binary data, but also
find optimal number of clusters in the data. The performance of the algorithm
was evaluated and compared with other algorithms using simulated datasets. It
outperformed all other clustering methods that were tested in the simulation
studies. It was also used to cluster real datasets arising from document
analysis, handwritten image analysis and cancer research. It successfully
divided a set of documents based on their topics, hand written images based on
different styles of writing digits and identified tissue and mutation
specificity of chemotherapy treatments.
| [
{
"version": "v1",
"created": "Tue, 8 Mar 2016 12:02:59 GMT"
}
] | 2016-03-09T00:00:00 | [
[
"Santra",
"Tapesh",
""
]
] | TITLE: A Bayesian non-parametric method for clustering high-dimensional binary
data
ABSTRACT: In many real life problems, objects are described by large number of binary
features. For instance, documents are characterized by presence or absence of
certain keywords; cancer patients are characterized by presence or absence of
certain mutations etc. In such cases, grouping together similar
objects/profiles based on such high dimensional binary features is desirable,
but challenging. Here, I present a Bayesian non parametric algorithm for
clustering high dimensional binary data. It uses a Dirichlet Process (DP)
mixture model and simulated annealing to not only cluster binary data, but also
find optimal number of clusters in the data. The performance of the algorithm
was evaluated and compared with other algorithms using simulated datasets. It
outperformed all other clustering methods that were tested in the simulation
studies. It was also used to cluster real datasets arising from document
analysis, handwritten image analysis and cancer research. It successfully
divided a set of documents based on their topics, hand written images based on
different styles of writing digits and identified tissue and mutation
specificity of chemotherapy treatments.
| no_new_dataset | 0.953923 |
1505.06236 | Le Lu | Amal Farag, Le Lu, Holger R. Roth, Jiamin Liu, Evrim Turkbey, Ronald
M. Summers | A Bottom-up Approach for Pancreas Segmentation using Cascaded
Superpixels and (Deep) Image Patch Labeling | 14 pages, 14 figures, 2 tables | null | null | null | cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | Robust automated organ segmentation is a prerequisite for computer-aided
diagnosis (CAD), quantitative imaging analysis and surgical assistance. For
high-variability organs such as the pancreas, previous approaches report
undesirably low accuracies. We present a bottom-up approach for pancreas
segmentation in abdominal CT scans that is based on a hierarchy of information
propagation by classifying image patches at different resolutions; and
cascading superpixels. There are four stages: 1) decomposing CT slice images as
a set of disjoint boundary-preserving superpixels; 2) computing pancreas class
probability maps via dense patch labeling; 3) classifying superpixels by
pooling both intensity and probability features to form empirical statistics in
cascaded random forest frameworks; and 4) simple connectivity based
post-processing. The dense image patch labeling are conducted by: efficient
random forest classifier on image histogram, location and texture features; and
more expensive (but with better specificity) deep convolutional neural network
classification on larger image windows (with more spatial contexts). Evaluation
of the approach is performed on a database of 80 manually segmented CT volumes
in six-fold cross-validation (CV). Our achieved results are comparable, or
better than the state-of-the-art methods (evaluated by
"leave-one-patient-out"), with Dice 70.7% and Jaccard 57.9%. The computational
efficiency has been drastically improved in the order of 6~8 minutes, comparing
with others of ~10 hours per case. Finally, we implement a multi-atlas label
fusion (MALF) approach for pancreas segmentation using the same datasets. Under
six-fold CV, our bottom-up segmentation method significantly outperforms its
MALF counterpart: (70.7 +/- 13.0%) versus (52.5 +/- 20.8%) in Dice. Deep CNN
patch labeling confidences offer more numerical stability, reflected by smaller
standard deviations.
| [
{
"version": "v1",
"created": "Fri, 22 May 2015 21:59:45 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Mar 2016 18:24:43 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Farag",
"Amal",
""
],
[
"Lu",
"Le",
""
],
[
"Roth",
"Holger R.",
""
],
[
"Liu",
"Jiamin",
""
],
[
"Turkbey",
"Evrim",
""
],
[
"Summers",
"Ronald M.",
""
]
] | TITLE: A Bottom-up Approach for Pancreas Segmentation using Cascaded
Superpixels and (Deep) Image Patch Labeling
ABSTRACT: Robust automated organ segmentation is a prerequisite for computer-aided
diagnosis (CAD), quantitative imaging analysis and surgical assistance. For
high-variability organs such as the pancreas, previous approaches report
undesirably low accuracies. We present a bottom-up approach for pancreas
segmentation in abdominal CT scans that is based on a hierarchy of information
propagation by classifying image patches at different resolutions; and
cascading superpixels. There are four stages: 1) decomposing CT slice images as
a set of disjoint boundary-preserving superpixels; 2) computing pancreas class
probability maps via dense patch labeling; 3) classifying superpixels by
pooling both intensity and probability features to form empirical statistics in
cascaded random forest frameworks; and 4) simple connectivity based
post-processing. The dense image patch labeling are conducted by: efficient
random forest classifier on image histogram, location and texture features; and
more expensive (but with better specificity) deep convolutional neural network
classification on larger image windows (with more spatial contexts). Evaluation
of the approach is performed on a database of 80 manually segmented CT volumes
in six-fold cross-validation (CV). Our achieved results are comparable, or
better than the state-of-the-art methods (evaluated by
"leave-one-patient-out"), with Dice 70.7% and Jaccard 57.9%. The computational
efficiency has been drastically improved in the order of 6~8 minutes, comparing
with others of ~10 hours per case. Finally, we implement a multi-atlas label
fusion (MALF) approach for pancreas segmentation using the same datasets. Under
six-fold CV, our bottom-up segmentation method significantly outperforms its
MALF counterpart: (70.7 +/- 13.0%) versus (52.5 +/- 20.8%) in Dice. Deep CNN
patch labeling confidences offer more numerical stability, reflected by smaller
standard deviations.
| no_new_dataset | 0.954095 |
1506.07285 | Richard Socher | Ankit Kumar and Ozan Irsoy and Peter Ondruska and Mohit Iyyer and
James Bradbury and Ishaan Gulrajani and Victor Zhong and Romain Paulus and
Richard Socher | Ask Me Anything: Dynamic Memory Networks for Natural Language Processing | null | null | null | null | cs.CL cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most tasks in natural language processing can be cast into question answering
(QA) problems over language input. We introduce the dynamic memory network
(DMN), a neural network architecture which processes input sequences and
questions, forms episodic memories, and generates relevant answers. Questions
trigger an iterative attention process which allows the model to condition its
attention on the inputs and the result of previous iterations. These results
are then reasoned over in a hierarchical recurrent sequence model to generate
answers. The DMN can be trained end-to-end and obtains state-of-the-art results
on several types of tasks and datasets: question answering (Facebook's bAbI
dataset), text classification for sentiment analysis (Stanford Sentiment
Treebank) and sequence modeling for part-of-speech tagging (WSJ-PTB). The
training for these different tasks relies exclusively on trained word vector
representations and input-question-answer triplets.
| [
{
"version": "v1",
"created": "Wed, 24 Jun 2015 08:27:02 GMT"
},
{
"version": "v2",
"created": "Fri, 24 Jul 2015 22:21:29 GMT"
},
{
"version": "v3",
"created": "Tue, 29 Sep 2015 05:02:29 GMT"
},
{
"version": "v4",
"created": "Tue, 9 Feb 2016 08:19:30 GMT"
},
{
"version": "v5",
"created": "Sat, 5 Mar 2016 20:18:55 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Kumar",
"Ankit",
""
],
[
"Irsoy",
"Ozan",
""
],
[
"Ondruska",
"Peter",
""
],
[
"Iyyer",
"Mohit",
""
],
[
"Bradbury",
"James",
""
],
[
"Gulrajani",
"Ishaan",
""
],
[
"Zhong",
"Victor",
""
],
[
"Paulus",
"Romain",
""
],
[
"Socher",
"Richard",
""
]
] | TITLE: Ask Me Anything: Dynamic Memory Networks for Natural Language Processing
ABSTRACT: Most tasks in natural language processing can be cast into question answering
(QA) problems over language input. We introduce the dynamic memory network
(DMN), a neural network architecture which processes input sequences and
questions, forms episodic memories, and generates relevant answers. Questions
trigger an iterative attention process which allows the model to condition its
attention on the inputs and the result of previous iterations. These results
are then reasoned over in a hierarchical recurrent sequence model to generate
answers. The DMN can be trained end-to-end and obtains state-of-the-art results
on several types of tasks and datasets: question answering (Facebook's bAbI
dataset), text classification for sentiment analysis (Stanford Sentiment
Treebank) and sequence modeling for part-of-speech tagging (WSJ-PTB). The
training for these different tasks relies exclusively on trained word vector
representations and input-question-answer triplets.
| no_new_dataset | 0.948537 |
1508.07630 | Vural Aksakalli | Vural Aksakalli and Milad Malekipirbazari | Feature Selection via Binary Simultaneous Perturbation Stochastic
Approximation | This is the Istanbul Sehir University Technical Report
#SHR-ISE-2016.01. A short version of this report has been accepted for
publication at Pattern Recognition Letters | null | null | SHR-ISE-2016.01 | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Feature selection (FS) has become an indispensable task in dealing with
today's highly complex pattern recognition problems with massive number of
features. In this study, we propose a new wrapper approach for FS based on
binary simultaneous perturbation stochastic approximation (BSPSA). This
pseudo-gradient descent stochastic algorithm starts with an initial feature
vector and moves toward the optimal feature vector via successive iterations.
In each iteration, the current feature vector's individual components are
perturbed simultaneously by random offsets from a qualified probability
distribution. We present computational experiments on datasets with numbers of
features ranging from a few dozens to thousands using three widely-used
classifiers as wrappers: nearest neighbor, decision tree, and linear support
vector machine. We compare our methodology against the full set of features as
well as a binary genetic algorithm and sequential FS methods using
cross-validated classification error rate and AUC as the performance criteria.
Our results indicate that features selected by BSPSA compare favorably to
alternative methods in general and BSPSA can yield superior feature sets for
datasets with tens of thousands of features by examining an extremely small
fraction of the solution space. We are not aware of any other wrapper FS
methods that are computationally feasible with good convergence properties for
such large datasets.
| [
{
"version": "v1",
"created": "Sun, 30 Aug 2015 20:03:53 GMT"
},
{
"version": "v2",
"created": "Thu, 14 Jan 2016 08:02:42 GMT"
},
{
"version": "v3",
"created": "Sat, 5 Mar 2016 19:42:14 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Aksakalli",
"Vural",
""
],
[
"Malekipirbazari",
"Milad",
""
]
] | TITLE: Feature Selection via Binary Simultaneous Perturbation Stochastic
Approximation
ABSTRACT: Feature selection (FS) has become an indispensable task in dealing with
today's highly complex pattern recognition problems with massive number of
features. In this study, we propose a new wrapper approach for FS based on
binary simultaneous perturbation stochastic approximation (BSPSA). This
pseudo-gradient descent stochastic algorithm starts with an initial feature
vector and moves toward the optimal feature vector via successive iterations.
In each iteration, the current feature vector's individual components are
perturbed simultaneously by random offsets from a qualified probability
distribution. We present computational experiments on datasets with numbers of
features ranging from a few dozens to thousands using three widely-used
classifiers as wrappers: nearest neighbor, decision tree, and linear support
vector machine. We compare our methodology against the full set of features as
well as a binary genetic algorithm and sequential FS methods using
cross-validated classification error rate and AUC as the performance criteria.
Our results indicate that features selected by BSPSA compare favorably to
alternative methods in general and BSPSA can yield superior feature sets for
datasets with tens of thousands of features by examining an extremely small
fraction of the solution space. We are not aware of any other wrapper FS
methods that are computationally feasible with good convergence properties for
such large datasets.
| no_new_dataset | 0.947914 |
1509.03611 | Ella Rabinovich | Ella Rabinovich, Shuly Wintner, Ofek Luis Lewinsohn | A Parallel Corpus of Translationese | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a set of bilingual English--French and English--German parallel
corpora in which the direction of translation is accurately and reliably
annotated. The corpora are diverse, consisting of parliamentary proceedings,
literary works, transcriptions of TED talks and political commentary. They will
be instrumental for research of translationese and its applications to (human
and machine) translation; specifically, they can be used for the task of
translationese identification, a research direction that enjoys a growing
interest in recent years. To validate the quality and reliability of the
corpora, we replicated previous results of supervised and unsupervised
identification of translationese, and further extended the experiments to
additional datasets and languages.
| [
{
"version": "v1",
"created": "Fri, 11 Sep 2015 19:07:49 GMT"
},
{
"version": "v2",
"created": "Sun, 6 Mar 2016 13:41:11 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Rabinovich",
"Ella",
""
],
[
"Wintner",
"Shuly",
""
],
[
"Lewinsohn",
"Ofek Luis",
""
]
] | TITLE: A Parallel Corpus of Translationese
ABSTRACT: We describe a set of bilingual English--French and English--German parallel
corpora in which the direction of translation is accurately and reliably
annotated. The corpora are diverse, consisting of parliamentary proceedings,
literary works, transcriptions of TED talks and political commentary. They will
be instrumental for research of translationese and its applications to (human
and machine) translation; specifically, they can be used for the task of
translationese identification, a research direction that enjoys a growing
interest in recent years. To validate the quality and reliability of the
corpora, we replicated previous results of supervised and unsupervised
identification of translationese, and further extended the experiments to
additional datasets and languages.
| no_new_dataset | 0.717507 |
1511.02258 | Ze Jia Zhang | Z. Zhang, K. Duraisamy, N. A. Gumerov | Efficient Multiscale Gaussian Process Regression using Hierarchical
Clustering | 22 pages, 9 figures. Preprint. Submitted to Machine Learning Mar.
2016 | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Standard Gaussian Process (GP) regression, a powerful machine learning tool,
is computationally expensive when it is applied to large datasets, and
potentially inaccurate when data points are sparsely distributed in a
high-dimensional feature space. To address these challenges, a new multiscale,
sparsified GP algorithm is formulated, with the goal of application to large
scientific computing datasets. In this approach, the data is partitioned into
clusters and the cluster centers are used to define a reduced training set,
resulting in an improvement over standard GPs in terms of training and
evaluation costs. Further, a hierarchical technique is used to adaptively map
the local covariance representation to the underlying sparsity of the feature
space, leading to improved prediction accuracy when the data distribution is
highly non-uniform. A theoretical investigation of the computational complexity
of the algorithm is presented. The efficacy of this method is then demonstrated
on smooth and discontinuous analytical functions and on data from a direct
numerical simulation of turbulent combustion.
| [
{
"version": "v1",
"created": "Fri, 6 Nov 2015 23:18:13 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Mar 2016 04:20:37 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Zhang",
"Z.",
""
],
[
"Duraisamy",
"K.",
""
],
[
"Gumerov",
"N. A.",
""
]
] | TITLE: Efficient Multiscale Gaussian Process Regression using Hierarchical
Clustering
ABSTRACT: Standard Gaussian Process (GP) regression, a powerful machine learning tool,
is computationally expensive when it is applied to large datasets, and
potentially inaccurate when data points are sparsely distributed in a
high-dimensional feature space. To address these challenges, a new multiscale,
sparsified GP algorithm is formulated, with the goal of application to large
scientific computing datasets. In this approach, the data is partitioned into
clusters and the cluster centers are used to define a reduced training set,
resulting in an improvement over standard GPs in terms of training and
evaluation costs. Further, a hierarchical technique is used to adaptively map
the local covariance representation to the underlying sparsity of the feature
space, leading to improved prediction accuracy when the data distribution is
highly non-uniform. A theoretical investigation of the computational complexity
of the algorithm is presented. The efficacy of this method is then demonstrated
on smooth and discontinuous analytical functions and on data from a direct
numerical simulation of turbulent combustion.
| no_new_dataset | 0.952574 |
1603.01716 | Balint Antal | B\'alint Antal | Classifier ensemble creation via false labelling | null | Knowledge-based Systems 89: pp. 278-287. (2015) | 10.1016/j.knosys.2015.07.009 | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a novel approach to classifier ensemble creation is presented.
While other ensemble creation techniques are based on careful selection of
existing classifiers or preprocessing of the data, the presented approach
automatically creates an optimal labelling for a number of classifiers, which
are then assigned to the original data instances and fed to classifiers. The
approach has been evaluated on high-dimensional biomedical datasets. The
results show that the approach outperformed individual approaches in all cases.
| [
{
"version": "v1",
"created": "Sat, 5 Mar 2016 12:01:00 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Antal",
"Bálint",
""
]
] | TITLE: Classifier ensemble creation via false labelling
ABSTRACT: In this paper, a novel approach to classifier ensemble creation is presented.
While other ensemble creation techniques are based on careful selection of
existing classifiers or preprocessing of the data, the presented approach
automatically creates an optimal labelling for a number of classifiers, which
are then assigned to the original data instances and fed to classifiers. The
approach has been evaluated on high-dimensional biomedical datasets. The
results show that the approach outperformed individual approaches in all cases.
| no_new_dataset | 0.955152 |
1603.01870 | Sougata Chaudhuri | Sougata Chaudhuri and Georgios Theocharous and Mohammad Ghavamzadeh | Personalized Advertisement Recommendation: A Ranking Approach to Address
the Ubiquitous Click Sparsity Problem | Under review | null | null | null | cs.LG cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problem of personalized advertisement recommendation (PAR),
which consist of a user visiting a system (website) and the system displaying
one of $K$ ads to the user. The system uses an internal ad recommendation
policy to map the user's profile (context) to one of the ads. The user either
clicks or ignores the ad and correspondingly, the system updates its
recommendation policy. PAR problem is usually tackled by scalable
\emph{contextual bandit} algorithms, where the policies are generally based on
classifiers. A practical problem in PAR is extreme click sparsity, due to very
few users actually clicking on ads. We systematically study the drawback of
using contextual bandit algorithms based on classifier-based policies, in face
of extreme click sparsity. We then suggest an alternate policy, based on
rankers, learnt by optimizing the Area Under the Curve (AUC) ranking loss,
which can significantly alleviate the problem of click sparsity. We conduct
extensive experiments on public datasets, as well as three industry proprietary
datasets, to illustrate the improvement in click-through-rate (CTR) obtained by
using the ranker-based policy over classifier-based policies.
| [
{
"version": "v1",
"created": "Sun, 6 Mar 2016 20:26:41 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Chaudhuri",
"Sougata",
""
],
[
"Theocharous",
"Georgios",
""
],
[
"Ghavamzadeh",
"Mohammad",
""
]
] | TITLE: Personalized Advertisement Recommendation: A Ranking Approach to Address
the Ubiquitous Click Sparsity Problem
ABSTRACT: We study the problem of personalized advertisement recommendation (PAR),
which consist of a user visiting a system (website) and the system displaying
one of $K$ ads to the user. The system uses an internal ad recommendation
policy to map the user's profile (context) to one of the ads. The user either
clicks or ignores the ad and correspondingly, the system updates its
recommendation policy. PAR problem is usually tackled by scalable
\emph{contextual bandit} algorithms, where the policies are generally based on
classifiers. A practical problem in PAR is extreme click sparsity, due to very
few users actually clicking on ads. We systematically study the drawback of
using contextual bandit algorithms based on classifier-based policies, in face
of extreme click sparsity. We then suggest an alternate policy, based on
rankers, learnt by optimizing the Area Under the Curve (AUC) ranking loss,
which can significantly alleviate the problem of click sparsity. We conduct
extensive experiments on public datasets, as well as three industry proprietary
datasets, to illustrate the improvement in click-through-rate (CTR) obtained by
using the ranker-based policy over classifier-based policies.
| no_new_dataset | 0.944944 |
1603.01929 | Junting Ye | Junting Ye, Santhosh Kumar, Leman Akoglu | Temporal Opinion Spam Detection by Multivariate Indicative Signals | 10 pages, 29 figures | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Online consumer reviews reflect the testimonials of real people, unlike
advertisements. As such, they have critical impact on potential consumers, and
indirectly on businesses. According to a Harvard study (Luca 2011), +1 rise in
star-rating increases revenue by 5-9%. Problematically, such financial
incentives have created a market for spammers to fabricate reviews, to unjustly
promote or demote businesses, activities known as opinion spam (Jindal and Liu
2008). A vast majority of existing work on this problem have formulations based
on static review data, with respective techniques operating in an offline
fashion. Spam campaigns, however, are intended to make most impact during their
course. Abnormal events triggered by spammers' activities could be masked in
the load of future events, which static analysis would fail to identify. In
this work, we approach the opinion spam problem with a temporal formulation.
Specifically, we monitor a list of carefully selected indicative signals of
opinion spam over time and design efficient techniques to both detect and
characterize abnormal events in real-time. Experiments on datasets from two
different review sites show that our approach is fast, effective, and practical
to be deployed in real-world systems.
| [
{
"version": "v1",
"created": "Mon, 7 Mar 2016 04:18:06 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Ye",
"Junting",
""
],
[
"Kumar",
"Santhosh",
""
],
[
"Akoglu",
"Leman",
""
]
] | TITLE: Temporal Opinion Spam Detection by Multivariate Indicative Signals
ABSTRACT: Online consumer reviews reflect the testimonials of real people, unlike
advertisements. As such, they have critical impact on potential consumers, and
indirectly on businesses. According to a Harvard study (Luca 2011), +1 rise in
star-rating increases revenue by 5-9%. Problematically, such financial
incentives have created a market for spammers to fabricate reviews, to unjustly
promote or demote businesses, activities known as opinion spam (Jindal and Liu
2008). A vast majority of existing work on this problem have formulations based
on static review data, with respective techniques operating in an offline
fashion. Spam campaigns, however, are intended to make most impact during their
course. Abnormal events triggered by spammers' activities could be masked in
the load of future events, which static analysis would fail to identify. In
this work, we approach the opinion spam problem with a temporal formulation.
Specifically, we monitor a list of carefully selected indicative signals of
opinion spam over time and design efficient techniques to both detect and
characterize abnormal events in real-time. Experiments on datasets from two
different review sites show that our approach is fast, effective, and practical
to be deployed in real-world systems.
| no_new_dataset | 0.937096 |
1603.02175 | Chunfeng Yang | Chunfeng Yang, Yipeng Zhou, Dah Ming Chiu | Who are Like-minded: Mining User Interest Similarity in Online Social
Networks | null | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we mine and learn to predict how similar a pair of users'
interests towards videos are, based on demographic (age, gender and location)
and social (friendship, interaction and group membership) information of these
users. We use the video access patterns of active users as ground truth (a form
of benchmark). We adopt tag-based user profiling to establish this ground
truth, and justify why it is used instead of video-based methods, or many
latent topic models such as LDA and Collaborative Filtering approaches. We then
show the effectiveness of the different demographic and social features, and
their combinations and derivatives, in predicting user interest similarity,
based on different machine-learning methods for combining multiple features. We
propose a hybrid tree-encoded linear model for combining the features, and show
that it out-performs other linear and treebased models. Our methods can be used
to predict user interest similarity when the ground-truth is not available,
e.g. for new users, or inactive users whose interests may have changed from old
access data, and is useful for video recommendation. Our study is based on a
rich dataset from Tencent, a popular service provider of social networks, video
services, and various other services in China.
| [
{
"version": "v1",
"created": "Mon, 7 Mar 2016 17:51:42 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Yang",
"Chunfeng",
""
],
[
"Zhou",
"Yipeng",
""
],
[
"Chiu",
"Dah Ming",
""
]
] | TITLE: Who are Like-minded: Mining User Interest Similarity in Online Social
Networks
ABSTRACT: In this paper, we mine and learn to predict how similar a pair of users'
interests towards videos are, based on demographic (age, gender and location)
and social (friendship, interaction and group membership) information of these
users. We use the video access patterns of active users as ground truth (a form
of benchmark). We adopt tag-based user profiling to establish this ground
truth, and justify why it is used instead of video-based methods, or many
latent topic models such as LDA and Collaborative Filtering approaches. We then
show the effectiveness of the different demographic and social features, and
their combinations and derivatives, in predicting user interest similarity,
based on different machine-learning methods for combining multiple features. We
propose a hybrid tree-encoded linear model for combining the features, and show
that it out-performs other linear and treebased models. Our methods can be used
to predict user interest similarity when the ground-truth is not available,
e.g. for new users, or inactive users whose interests may have changed from old
access data, and is useful for video recommendation. Our study is based on a
rich dataset from Tencent, a popular service provider of social networks, video
services, and various other services in China.
| no_new_dataset | 0.944485 |
1603.02211 | Rajesh Kumar | Rajesh Kumar, Vir V Phoha, and Rahul Raina | Authenticating users through their arm movement patterns | null | null | null | null | cs.CV cs.CR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we propose four continuous authentication designs by using the
characteristics of arm movements while individuals walk. The first design uses
acceleration of arms captured by a smartwatch's accelerometer sensor, the
second design uses the rotation of arms captured by a smartwatch's gyroscope
sensor, third uses the fusion of both acceleration and rotation at the
feature-level and fourth uses the fusion at score-level. Each of these designs
is implemented by using four classifiers, namely, k nearest neighbors (k-NN)
with Euclidean distance, Logistic Regression, Multilayer Perceptrons, and
Random Forest resulting in a total of sixteen authentication mechanisms. These
authentication mechanisms are tested under three different environments, namely
an intra-session, inter-session on a dataset of 40 users and an inter-phase on
a dataset of 12 users. The sessions of data collection were separated by at
least ten minutes, whereas the phases of data collection were separated by at
least three months. Under the intra-session environment, all of the twelve
authentication mechanisms achieve a mean dynamic false accept rate (DFAR) of 0%
and dynamic false reject rate (DFRR) of 0%. For the inter-session environment,
feature level fusion-based design with classifier k-NN achieves the best error
rates that are a mean DFAR of 2.2% and DFRR of 4.2%. The DFAR and DFRR
increased from 5.68% and 4.23% to 15.03% and 14.62% respectively when feature
level fusion-based design with classifier k-NN was tested under the inter-phase
environment on a dataset of 12 users.
| [
{
"version": "v1",
"created": "Mon, 7 Mar 2016 19:15:39 GMT"
}
] | 2016-03-08T00:00:00 | [
[
"Kumar",
"Rajesh",
""
],
[
"Phoha",
"Vir V",
""
],
[
"Raina",
"Rahul",
""
]
] | TITLE: Authenticating users through their arm movement patterns
ABSTRACT: In this paper, we propose four continuous authentication designs by using the
characteristics of arm movements while individuals walk. The first design uses
acceleration of arms captured by a smartwatch's accelerometer sensor, the
second design uses the rotation of arms captured by a smartwatch's gyroscope
sensor, third uses the fusion of both acceleration and rotation at the
feature-level and fourth uses the fusion at score-level. Each of these designs
is implemented by using four classifiers, namely, k nearest neighbors (k-NN)
with Euclidean distance, Logistic Regression, Multilayer Perceptrons, and
Random Forest resulting in a total of sixteen authentication mechanisms. These
authentication mechanisms are tested under three different environments, namely
an intra-session, inter-session on a dataset of 40 users and an inter-phase on
a dataset of 12 users. The sessions of data collection were separated by at
least ten minutes, whereas the phases of data collection were separated by at
least three months. Under the intra-session environment, all of the twelve
authentication mechanisms achieve a mean dynamic false accept rate (DFAR) of 0%
and dynamic false reject rate (DFRR) of 0%. For the inter-session environment,
feature level fusion-based design with classifier k-NN achieves the best error
rates that are a mean DFAR of 2.2% and DFRR of 4.2%. The DFAR and DFRR
increased from 5.68% and 4.23% to 15.03% and 14.62% respectively when feature
level fusion-based design with classifier k-NN was tested under the inter-phase
environment on a dataset of 12 users.
| no_new_dataset | 0.947284 |
1504.06103 | Tomas Vojir | Tomas Vojir, Jiri Matas, Jana Noskova | Online Adaptive Hidden Markov Model for Multi-Tracker Fusion | 27 pages, 9 figures, submitted to CVIU journal | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a novel method for visual object tracking called
HMMTxD. The method fuses observations from complementary out-of-the box
trackers and a detector by utilizing a hidden Markov model whose latent states
correspond to a binary vector expressing the failure of individual trackers.
The Markov model is trained in an unsupervised way, relying on an online
learned detector to provide a source of tracker-independent information for a
modified Baum- Welch algorithm that updates the model w.r.t. the partially
annotated data.
We show the effectiveness of the proposed method on combination of two and
three tracking algorithms. The performance of HMMTxD is evaluated on two
standard benchmarks (CVPR2013 and VOT) and on a rich collection of 77 publicly
available sequences. The HMMTxD outperforms the state-of-the-art, often
significantly, on all datasets in almost all criteria.
| [
{
"version": "v1",
"created": "Thu, 23 Apr 2015 09:34:59 GMT"
},
{
"version": "v2",
"created": "Fri, 4 Mar 2016 14:52:18 GMT"
}
] | 2016-03-07T00:00:00 | [
[
"Vojir",
"Tomas",
""
],
[
"Matas",
"Jiri",
""
],
[
"Noskova",
"Jana",
""
]
] | TITLE: Online Adaptive Hidden Markov Model for Multi-Tracker Fusion
ABSTRACT: In this paper, we propose a novel method for visual object tracking called
HMMTxD. The method fuses observations from complementary out-of-the box
trackers and a detector by utilizing a hidden Markov model whose latent states
correspond to a binary vector expressing the failure of individual trackers.
The Markov model is trained in an unsupervised way, relying on an online
learned detector to provide a source of tracker-independent information for a
modified Baum- Welch algorithm that updates the model w.r.t. the partially
annotated data.
We show the effectiveness of the proposed method on combination of two and
three tracking algorithms. The performance of HMMTxD is evaluated on two
standard benchmarks (CVPR2013 and VOT) and on a rich collection of 77 publicly
available sequences. The HMMTxD outperforms the state-of-the-art, often
significantly, on all datasets in almost all criteria.
| no_new_dataset | 0.945248 |
1506.08316 | Jianshu Chao | Jianshu Chao and Eckehard Steinbach | Keypoint Encoding for Improved Feature Extraction from Compressed Video
at Low Bitrates | null | null | null | null | cs.MM cs.CV cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In many mobile visual analysis applications, compressed video is transmitted
over a communication network and analyzed by a server. Typical processing steps
performed at the server include keypoint detection, descriptor calculation, and
feature matching. Video compression has been shown to have an adverse effect on
feature-matching performance. The negative impact of compression can be reduced
by using the keypoints extracted from the uncompressed video to calculate
descriptors from the compressed video. Based on this observation, we propose to
provide these keypoints to the server as side information and to extract only
the descriptors from the compressed video. First, we introduce four different
frame types for keypoint encoding to address different types of changes in
video content. These frame types represent a new scene, the same scene, a
slowly changing scene, or a rapidly moving scene and are determined by
comparing features between successive video frames. Then, we propose Intra,
Skip and Inter modes of encoding the keypoints for different frame types. For
example, keypoints for new scenes are encoded using the Intra mode, and
keypoints for unchanged scenes are skipped. As a result, the bitrate of the
side information related to keypoint encoding is significantly reduced.
Finally, we present pairwise matching and image retrieval experiments conducted
to evaluate the performance of the proposed approach using the Stanford mobile
augmented reality dataset and 720p format videos. The results show that the
proposed approach offers significantly improved feature matching and image
retrieval performance at a given bitrate.
| [
{
"version": "v1",
"created": "Sat, 27 Jun 2015 17:33:34 GMT"
},
{
"version": "v2",
"created": "Fri, 4 Mar 2016 16:43:42 GMT"
}
] | 2016-03-07T00:00:00 | [
[
"Chao",
"Jianshu",
""
],
[
"Steinbach",
"Eckehard",
""
]
] | TITLE: Keypoint Encoding for Improved Feature Extraction from Compressed Video
at Low Bitrates
ABSTRACT: In many mobile visual analysis applications, compressed video is transmitted
over a communication network and analyzed by a server. Typical processing steps
performed at the server include keypoint detection, descriptor calculation, and
feature matching. Video compression has been shown to have an adverse effect on
feature-matching performance. The negative impact of compression can be reduced
by using the keypoints extracted from the uncompressed video to calculate
descriptors from the compressed video. Based on this observation, we propose to
provide these keypoints to the server as side information and to extract only
the descriptors from the compressed video. First, we introduce four different
frame types for keypoint encoding to address different types of changes in
video content. These frame types represent a new scene, the same scene, a
slowly changing scene, or a rapidly moving scene and are determined by
comparing features between successive video frames. Then, we propose Intra,
Skip and Inter modes of encoding the keypoints for different frame types. For
example, keypoints for new scenes are encoded using the Intra mode, and
keypoints for unchanged scenes are skipped. As a result, the bitrate of the
side information related to keypoint encoding is significantly reduced.
Finally, we present pairwise matching and image retrieval experiments conducted
to evaluate the performance of the proposed approach using the Stanford mobile
augmented reality dataset and 720p format videos. The results show that the
proposed approach offers significantly improved feature matching and image
retrieval performance at a given bitrate.
| no_new_dataset | 0.954647 |
1509.02805 | Teng Qiu | Teng Qiu, Yongjie Li | Clustering by Hierarchical Nearest Neighbor Descent (H-NND) | 19 pages, 9 figures | null | null | null | stat.ML cs.CV cs.LG stat.ME | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Previously in 2014, we proposed the Nearest Descent (ND) method, capable of
generating an efficient Graph, called the in-tree (IT). Due to some beautiful
and effective features, this IT structure proves well suited for data
clustering. Although there exist some redundant edges in IT, they usually have
salient features and thus it is not hard to remove them.
Subsequently, in order to prevent the seemingly redundant edges from
occurring, we proposed the Nearest Neighbor Descent (NND) by adding the
"Neighborhood" constraint on ND. Consequently, clusters automatically emerged,
without the additional requirement of removing the redundant edges. However,
NND proved still not perfect, since it brought in a new yet worse problem, the
"over-partitioning" problem.
Now, in this paper, we propose a method, called the Hierarchical Nearest
Neighbor Descent (H-NND), which overcomes the over-partitioning problem of NND
via using the hierarchical strategy. Specifically, H-NND uses ND to effectively
merge the over-segmented sub-graphs or clusters that NND produces. Like ND,
H-NND also generates the IT structure, in which the redundant edges once again
appear. This seemingly comes back to the situation that ND faces. However,
compared with ND, the redundant edges in the IT structure generated by H-NND
generally become more salient, thus being much easier and more reliable to be
identified even by the simplest edge-removing method which takes the edge
length as the only measure. In other words, the IT structure constructed by
H-NND becomes more fitted for data clustering. We prove this on several
clustering datasets of varying shapes, dimensions and attributes. Besides,
compared with ND, H-NND generally takes less computation time to construct the
IT data structure for the input data.
| [
{
"version": "v1",
"created": "Wed, 9 Sep 2015 15:15:44 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Sep 2015 15:43:25 GMT"
},
{
"version": "v3",
"created": "Fri, 4 Mar 2016 15:50:58 GMT"
}
] | 2016-03-07T00:00:00 | [
[
"Qiu",
"Teng",
""
],
[
"Li",
"Yongjie",
""
]
] | TITLE: Clustering by Hierarchical Nearest Neighbor Descent (H-NND)
ABSTRACT: Previously in 2014, we proposed the Nearest Descent (ND) method, capable of
generating an efficient Graph, called the in-tree (IT). Due to some beautiful
and effective features, this IT structure proves well suited for data
clustering. Although there exist some redundant edges in IT, they usually have
salient features and thus it is not hard to remove them.
Subsequently, in order to prevent the seemingly redundant edges from
occurring, we proposed the Nearest Neighbor Descent (NND) by adding the
"Neighborhood" constraint on ND. Consequently, clusters automatically emerged,
without the additional requirement of removing the redundant edges. However,
NND proved still not perfect, since it brought in a new yet worse problem, the
"over-partitioning" problem.
Now, in this paper, we propose a method, called the Hierarchical Nearest
Neighbor Descent (H-NND), which overcomes the over-partitioning problem of NND
via using the hierarchical strategy. Specifically, H-NND uses ND to effectively
merge the over-segmented sub-graphs or clusters that NND produces. Like ND,
H-NND also generates the IT structure, in which the redundant edges once again
appear. This seemingly comes back to the situation that ND faces. However,
compared with ND, the redundant edges in the IT structure generated by H-NND
generally become more salient, thus being much easier and more reliable to be
identified even by the simplest edge-removing method which takes the edge
length as the only measure. In other words, the IT structure constructed by
H-NND becomes more fitted for data clustering. We prove this on several
clustering datasets of varying shapes, dimensions and attributes. Besides,
compared with ND, H-NND generally takes less computation time to construct the
IT data structure for the input data.
| no_new_dataset | 0.948442 |
1511.08198 | John Wieting | John Wieting, Mohit Bansal, Kevin Gimpel, Karen Livescu | Towards Universal Paraphrastic Sentence Embeddings | Published as a conference paper at ICLR 2016 | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of learning general-purpose, paraphrastic sentence
embeddings based on supervision from the Paraphrase Database (Ganitkevitch et
al., 2013). We compare six compositional architectures, evaluating them on
annotated textual similarity datasets drawn both from the same distribution as
the training data and from a wide range of other domains. We find that the most
complex architectures, such as long short-term memory (LSTM) recurrent neural
networks, perform best on the in-domain data. However, in out-of-domain
scenarios, simple architectures such as word averaging vastly outperform LSTMs.
Our simplest averaging model is even competitive with systems tuned for the
particular tasks while also being extremely efficient and easy to use.
In order to better understand how these architectures compare, we conduct
further experiments on three supervised NLP tasks: sentence similarity,
entailment, and sentiment classification. We again find that the word averaging
models perform well for sentence similarity and entailment, outperforming
LSTMs. However, on sentiment classification, we find that the LSTM performs
very strongly-even recording new state-of-the-art performance on the Stanford
Sentiment Treebank.
We then demonstrate how to combine our pretrained sentence embeddings with
these supervised tasks, using them both as a prior and as a black box feature
extractor. This leads to performance rivaling the state of the art on the SICK
similarity and entailment tasks. We release all of our resources to the
research community with the hope that they can serve as the new baseline for
further work on universal sentence embeddings.
| [
{
"version": "v1",
"created": "Wed, 25 Nov 2015 20:52:15 GMT"
},
{
"version": "v2",
"created": "Tue, 12 Jan 2016 20:59:39 GMT"
},
{
"version": "v3",
"created": "Fri, 4 Mar 2016 20:54:30 GMT"
}
] | 2016-03-07T00:00:00 | [
[
"Wieting",
"John",
""
],
[
"Bansal",
"Mohit",
""
],
[
"Gimpel",
"Kevin",
""
],
[
"Livescu",
"Karen",
""
]
] | TITLE: Towards Universal Paraphrastic Sentence Embeddings
ABSTRACT: We consider the problem of learning general-purpose, paraphrastic sentence
embeddings based on supervision from the Paraphrase Database (Ganitkevitch et
al., 2013). We compare six compositional architectures, evaluating them on
annotated textual similarity datasets drawn both from the same distribution as
the training data and from a wide range of other domains. We find that the most
complex architectures, such as long short-term memory (LSTM) recurrent neural
networks, perform best on the in-domain data. However, in out-of-domain
scenarios, simple architectures such as word averaging vastly outperform LSTMs.
Our simplest averaging model is even competitive with systems tuned for the
particular tasks while also being extremely efficient and easy to use.
In order to better understand how these architectures compare, we conduct
further experiments on three supervised NLP tasks: sentence similarity,
entailment, and sentiment classification. We again find that the word averaging
models perform well for sentence similarity and entailment, outperforming
LSTMs. However, on sentiment classification, we find that the LSTM performs
very strongly-even recording new state-of-the-art performance on the Stanford
Sentiment Treebank.
We then demonstrate how to combine our pretrained sentence embeddings with
these supervised tasks, using them both as a prior and as a black box feature
extractor. This leads to performance rivaling the state of the art on the SICK
similarity and entailment tasks. We release all of our resources to the
research community with the hope that they can serve as the new baseline for
further work on universal sentence embeddings.
| no_new_dataset | 0.944587 |
1603.01336 | Sabir Ribas | Sabir Ribas, Alberto Ueda, Rodrygo L. T. Santos, Berthier
Ribeiro-Neto, Nivio Ziviani | Simplified Relative Citation Ratio for Static Paper Ranking: UFMG/LATIN
at WSDM Cup 2016 | WSDM Cup. The 9th ACM International Conference on Web Search and Data
Mining San Francisco, California, USA. February 22-25, 2016 | null | null | null | cs.IR cs.DL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Static rankings of papers play a key role in the academic search setting.
Many features are commonly used in the literature to produce such rankings,
some examples are citation-based metrics, distinct applications of PageRank,
among others. More recently, learning to rank techniques have been successfully
applied to combine sets of features producing effective results. In this work,
we propose the metric S-RCR, which is a simplified version of a metric called
Relative Citation Ratio --- both based on the idea of a co-citation network.
When compared to the classical version, our simplification S-RCR leads to
improved efficiency with a reasonable effectiveness. We use S-RCR to rank over
120 million papers in the Microsoft Academic Graph dataset. By using this
single feature, which has no parameters and does not need to be tuned, our team
was able to reach the 3rd position in the first phase of the WSDM Cup 2016.
| [
{
"version": "v1",
"created": "Fri, 4 Mar 2016 03:00:46 GMT"
}
] | 2016-03-07T00:00:00 | [
[
"Ribas",
"Sabir",
""
],
[
"Ueda",
"Alberto",
""
],
[
"Santos",
"Rodrygo L. T.",
""
],
[
"Ribeiro-Neto",
"Berthier",
""
],
[
"Ziviani",
"Nivio",
""
]
] | TITLE: Simplified Relative Citation Ratio for Static Paper Ranking: UFMG/LATIN
at WSDM Cup 2016
ABSTRACT: Static rankings of papers play a key role in the academic search setting.
Many features are commonly used in the literature to produce such rankings,
some examples are citation-based metrics, distinct applications of PageRank,
among others. More recently, learning to rank techniques have been successfully
applied to combine sets of features producing effective results. In this work,
we propose the metric S-RCR, which is a simplified version of a metric called
Relative Citation Ratio --- both based on the idea of a co-citation network.
When compared to the classical version, our simplification S-RCR leads to
improved efficiency with a reasonable effectiveness. We use S-RCR to rank over
120 million papers in the Microsoft Academic Graph dataset. By using this
single feature, which has no parameters and does not need to be tuned, our team
was able to reach the 3rd position in the first phase of the WSDM Cup 2016.
| no_new_dataset | 0.9434 |
1603.01417 | Richard Socher | Caiming Xiong, Stephen Merity, Richard Socher | Dynamic Memory Networks for Visual and Textual Question Answering | null | null | null | null | cs.NE cs.CL cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural network architectures with memory and attention mechanisms exhibit
certain reasoning capabilities required for question answering. One such
architecture, the dynamic memory network (DMN), obtained high accuracy on a
variety of language tasks. However, it was not shown whether the architecture
achieves strong results for question answering when supporting facts are not
marked during training or whether it could be applied to other modalities such
as images. Based on an analysis of the DMN, we propose several improvements to
its memory and input modules. Together with these changes we introduce a novel
input module for images in order to be able to answer visual questions. Our new
DMN+ model improves the state of the art on both the Visual Question Answering
dataset and the \babi-10k text question-answering dataset without supporting
fact supervision.
| [
{
"version": "v1",
"created": "Fri, 4 Mar 2016 10:40:28 GMT"
}
] | 2016-03-07T00:00:00 | [
[
"Xiong",
"Caiming",
""
],
[
"Merity",
"Stephen",
""
],
[
"Socher",
"Richard",
""
]
] | TITLE: Dynamic Memory Networks for Visual and Textual Question Answering
ABSTRACT: Neural network architectures with memory and attention mechanisms exhibit
certain reasoning capabilities required for question answering. One such
architecture, the dynamic memory network (DMN), obtained high accuracy on a
variety of language tasks. However, it was not shown whether the architecture
achieves strong results for question answering when supporting facts are not
marked during training or whether it could be applied to other modalities such
as images. Based on an analysis of the DMN, we propose several improvements to
its memory and input modules. Together with these changes we introduce a novel
input module for images in order to be able to answer visual questions. Our new
DMN+ model improves the state of the art on both the Visual Question Answering
dataset and the \babi-10k text question-answering dataset without supporting
fact supervision.
| no_new_dataset | 0.947575 |
1603.00892 | Nikola Mrk\v{s}i\'c | Nikola Mrk\v{s}i\'c and Diarmuid \'O S\'eaghdha and Blaise Thomson and
Milica Ga\v{s}i\'c and Lina Rojas-Barahona and Pei-Hao Su and David Vandyke
and Tsung-Hsien Wen and Steve Young | Counter-fitting Word Vectors to Linguistic Constraints | Paper accepted for the 15th Annual Conference of the North American
Chapter of the Association for Computational Linguistics (NAACL 2016) | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present a novel counter-fitting method which injects
antonymy and synonymy constraints into vector space representations in order to
improve the vectors' capability for judging semantic similarity. Applying this
method to publicly available pre-trained word vectors leads to a new state of
the art performance on the SimLex-999 dataset. We also show how the method can
be used to tailor the word vector space for the downstream task of dialogue
state tracking, resulting in robust improvements across different dialogue
domains.
| [
{
"version": "v1",
"created": "Wed, 2 Mar 2016 21:19:36 GMT"
}
] | 2016-03-04T00:00:00 | [
[
"Mrkšić",
"Nikola",
""
],
[
"Séaghdha",
"Diarmuid Ó",
""
],
[
"Thomson",
"Blaise",
""
],
[
"Gašić",
"Milica",
""
],
[
"Rojas-Barahona",
"Lina",
""
],
[
"Su",
"Pei-Hao",
""
],
[
"Vandyke",
"David",
""
],
[
"Wen",
"Tsung-Hsien",
""
],
[
"Young",
"Steve",
""
]
] | TITLE: Counter-fitting Word Vectors to Linguistic Constraints
ABSTRACT: In this work, we present a novel counter-fitting method which injects
antonymy and synonymy constraints into vector space representations in order to
improve the vectors' capability for judging semantic similarity. Applying this
method to publicly available pre-trained word vectors leads to a new state of
the art performance on the SimLex-999 dataset. We also show how the method can
be used to tailor the word vector space for the downstream task of dialogue
state tracking, resulting in robust improvements across different dialogue
domains.
| no_new_dataset | 0.952662 |
1603.01046 | Teemu Helenius | Teemu Helenius and Samuli Siltanen | Photographic dataset: random peppercorns | null | null | null | null | physics.data-an cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This is a photographic dataset collected for testing image processing
algorithms. The idea is to have sets of different but statistically similar
images. In this work the images show randomly distributed peppercorns. The
dataset is made available at www.fips.fi/photographic_dataset.php .
| [
{
"version": "v1",
"created": "Thu, 3 Mar 2016 10:24:07 GMT"
}
] | 2016-03-04T00:00:00 | [
[
"Helenius",
"Teemu",
""
],
[
"Siltanen",
"Samuli",
""
]
] | TITLE: Photographic dataset: random peppercorns
ABSTRACT: This is a photographic dataset collected for testing image processing
algorithms. The idea is to have sets of different but statistically similar
images. In this work the images show randomly distributed peppercorns. The
dataset is made available at www.fips.fi/photographic_dataset.php .
| new_dataset | 0.953579 |
1603.01232 | Tsung-Hsien Wen | Tsung-Hsien Wen, Milica Gasic, Nikola Mrksic, Lina M. Rojas-Barahona,
Pei-Hao Su, David Vandyke, Steve Young | Multi-domain Neural Network Language Generation for Spoken Dialogue
Systems | Accepted as a long paper in NAACL-HLT 2016 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Moving from limited-domain natural language generation (NLG) to open domain
is difficult because the number of semantic input combinations grows
exponentially with the number of domains. Therefore, it is important to
leverage existing resources and exploit similarities between domains to
facilitate domain adaptation. In this paper, we propose a procedure to train
multi-domain, Recurrent Neural Network-based (RNN) language generators via
multiple adaptation steps. In this procedure, a model is first trained on
counterfeited data synthesised from an out-of-domain dataset, and then fine
tuned on a small set of in-domain utterances with a discriminative objective
function. Corpus-based evaluation results show that the proposed procedure can
achieve competitive performance in terms of BLEU score and slot error rate
while significantly reducing the data needed to train generators in new, unseen
domains. In subjective testing, human judges confirm that the procedure greatly
improves generator performance when only a small amount of data is available in
the domain.
| [
{
"version": "v1",
"created": "Thu, 3 Mar 2016 19:49:32 GMT"
}
] | 2016-03-04T00:00:00 | [
[
"Wen",
"Tsung-Hsien",
""
],
[
"Gasic",
"Milica",
""
],
[
"Mrksic",
"Nikola",
""
],
[
"Rojas-Barahona",
"Lina M.",
""
],
[
"Su",
"Pei-Hao",
""
],
[
"Vandyke",
"David",
""
],
[
"Young",
"Steve",
""
]
] | TITLE: Multi-domain Neural Network Language Generation for Spoken Dialogue
Systems
ABSTRACT: Moving from limited-domain natural language generation (NLG) to open domain
is difficult because the number of semantic input combinations grows
exponentially with the number of domains. Therefore, it is important to
leverage existing resources and exploit similarities between domains to
facilitate domain adaptation. In this paper, we propose a procedure to train
multi-domain, Recurrent Neural Network-based (RNN) language generators via
multiple adaptation steps. In this procedure, a model is first trained on
counterfeited data synthesised from an out-of-domain dataset, and then fine
tuned on a small set of in-domain utterances with a discriminative objective
function. Corpus-based evaluation results show that the proposed procedure can
achieve competitive performance in terms of BLEU score and slot error rate
while significantly reducing the data needed to train generators in new, unseen
domains. In subjective testing, human judges confirm that the procedure greatly
improves generator performance when only a small amount of data is available in
the domain.
| no_new_dataset | 0.951414 |
1603.01250 | Yani Ioannou | Yani Ioannou, Duncan Robertson, Darko Zikic, Peter Kontschieder, Jamie
Shotton, Matthew Brown, and Antonio Criminisi | Decision Forests, Convolutional Networks and the Models in-Between | Microsoft Research Technical Report | null | null | MSR-TR-2015-58 | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates the connections between two state of the art
classifiers: decision forests (DFs, including decision jungles) and
convolutional neural networks (CNNs). Decision forests are computationally
efficient thanks to their conditional computation property (computation is
confined to only a small region of the tree, the nodes along a single branch).
CNNs achieve state of the art accuracy, thanks to their representation learning
capabilities. We present a systematic analysis of how to fuse conditional
computation with representation learning and achieve a continuum of hybrid
models with different ratios of accuracy vs. efficiency. We call this new
family of hybrid models conditional networks. Conditional networks can be
thought of as: i) decision trees augmented with data transformation operators,
or ii) CNNs, with block-diagonal sparse weight matrices, and explicit data
routing functions. Experimental validation is performed on the common task of
image classification on both the CIFAR and Imagenet datasets. Compared to state
of the art CNNs, our hybrid models yield the same accuracy with a fraction of
the compute cost and much smaller number of parameters.
| [
{
"version": "v1",
"created": "Thu, 3 Mar 2016 20:41:47 GMT"
}
] | 2016-03-04T00:00:00 | [
[
"Ioannou",
"Yani",
""
],
[
"Robertson",
"Duncan",
""
],
[
"Zikic",
"Darko",
""
],
[
"Kontschieder",
"Peter",
""
],
[
"Shotton",
"Jamie",
""
],
[
"Brown",
"Matthew",
""
],
[
"Criminisi",
"Antonio",
""
]
] | TITLE: Decision Forests, Convolutional Networks and the Models in-Between
ABSTRACT: This paper investigates the connections between two state of the art
classifiers: decision forests (DFs, including decision jungles) and
convolutional neural networks (CNNs). Decision forests are computationally
efficient thanks to their conditional computation property (computation is
confined to only a small region of the tree, the nodes along a single branch).
CNNs achieve state of the art accuracy, thanks to their representation learning
capabilities. We present a systematic analysis of how to fuse conditional
computation with representation learning and achieve a continuum of hybrid
models with different ratios of accuracy vs. efficiency. We call this new
family of hybrid models conditional networks. Conditional networks can be
thought of as: i) decision trees augmented with data transformation operators,
or ii) CNNs, with block-diagonal sparse weight matrices, and explicit data
routing functions. Experimental validation is performed on the common task of
image classification on both the CIFAR and Imagenet datasets. Compared to state
of the art CNNs, our hybrid models yield the same accuracy with a fraction of
the compute cost and much smaller number of parameters.
| no_new_dataset | 0.952131 |
1506.05011 | Theofanis Karaletsos | Theofanis Karaletsos, Serge Belongie, Gunnar R\"atsch | Bayesian representation learning with oracle constraints | 16 pages, publishes in ICLR 16 | null | null | null | stat.ML cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Representation learning systems typically rely on massive amounts of labeled
data in order to be trained to high accuracy. Recently, high-dimensional
parametric models like neural networks have succeeded in building rich
representations using either compressive, reconstructive or supervised
criteria. However, the semantic structure inherent in observations is
oftentimes lost in the process. Human perception excels at understanding
semantics but cannot always be expressed in terms of labels. Thus,
\emph{oracles} or \emph{human-in-the-loop systems}, for example crowdsourcing,
are often employed to generate similarity constraints using an implicit
similarity function encoded in human perception. In this work we propose to
combine \emph{generative unsupervised feature learning} with a
\emph{probabilistic treatment of oracle information like triplets} in order to
transfer implicit privileged oracle knowledge into explicit nonlinear Bayesian
latent factor models of the observations. We use a fast variational algorithm
to learn the joint model and demonstrate applicability to a well-known image
dataset. We show how implicit triplet information can provide rich information
to learn representations that outperform previous metric learning approaches as
well as generative models without this side-information in a variety of
predictive tasks. In addition, we illustrate that the proposed approach
compartmentalizes the latent spaces semantically which allows interpretation of
the latent variables.
| [
{
"version": "v1",
"created": "Tue, 16 Jun 2015 15:54:59 GMT"
},
{
"version": "v2",
"created": "Sat, 21 Nov 2015 05:24:01 GMT"
},
{
"version": "v3",
"created": "Fri, 8 Jan 2016 04:47:21 GMT"
},
{
"version": "v4",
"created": "Tue, 1 Mar 2016 23:36:04 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Karaletsos",
"Theofanis",
""
],
[
"Belongie",
"Serge",
""
],
[
"Rätsch",
"Gunnar",
""
]
] | TITLE: Bayesian representation learning with oracle constraints
ABSTRACT: Representation learning systems typically rely on massive amounts of labeled
data in order to be trained to high accuracy. Recently, high-dimensional
parametric models like neural networks have succeeded in building rich
representations using either compressive, reconstructive or supervised
criteria. However, the semantic structure inherent in observations is
oftentimes lost in the process. Human perception excels at understanding
semantics but cannot always be expressed in terms of labels. Thus,
\emph{oracles} or \emph{human-in-the-loop systems}, for example crowdsourcing,
are often employed to generate similarity constraints using an implicit
similarity function encoded in human perception. In this work we propose to
combine \emph{generative unsupervised feature learning} with a
\emph{probabilistic treatment of oracle information like triplets} in order to
transfer implicit privileged oracle knowledge into explicit nonlinear Bayesian
latent factor models of the observations. We use a fast variational algorithm
to learn the joint model and demonstrate applicability to a well-known image
dataset. We show how implicit triplet information can provide rich information
to learn representations that outperform previous metric learning approaches as
well as generative models without this side-information in a variety of
predictive tasks. In addition, we illustrate that the proposed approach
compartmentalizes the latent spaces semantically which allows interpretation of
the latent variables.
| no_new_dataset | 0.946646 |
1508.01722 | Jun-Cheng Chen | Jun-Cheng Chen and Vishal M. Patel and Rama Chellappa | Unconstrained Face Verification using Deep CNN Features | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present an algorithm for unconstrained face verification
based on deep convolutional features and evaluate it on the newly released
IARPA Janus Benchmark A (IJB-A) dataset. The IJB-A dataset includes real-world
unconstrained faces from 500 subjects with full pose and illumination
variations which are much harder than the traditional Labeled Face in the Wild
(LFW) and Youtube Face (YTF) datasets. The deep convolutional neural network
(DCNN) is trained using the CASIA-WebFace dataset. Extensive experiments on the
IJB-A dataset are provided.
| [
{
"version": "v1",
"created": "Fri, 7 Aug 2015 15:21:19 GMT"
},
{
"version": "v2",
"created": "Wed, 2 Mar 2016 19:41:42 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Chen",
"Jun-Cheng",
""
],
[
"Patel",
"Vishal M.",
""
],
[
"Chellappa",
"Rama",
""
]
] | TITLE: Unconstrained Face Verification using Deep CNN Features
ABSTRACT: In this paper, we present an algorithm for unconstrained face verification
based on deep convolutional features and evaluate it on the newly released
IARPA Janus Benchmark A (IJB-A) dataset. The IJB-A dataset includes real-world
unconstrained faces from 500 subjects with full pose and illumination
variations which are much harder than the traditional Labeled Face in the Wild
(LFW) and Youtube Face (YTF) datasets. The deep convolutional neural network
(DCNN) is trained using the CASIA-WebFace dataset. Extensive experiments on the
IJB-A dataset are provided.
| new_dataset | 0.643329 |
1511.05939 | Oren Rippel | Oren Rippel, Manohar Paluri, Piotr Dollar, Lubomir Bourdev | Metric Learning with Adaptive Density Discrimination | ICLR 2016 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Distance metric learning (DML) approaches learn a transformation to a
representation space where distance is in correspondence with a predefined
notion of similarity. While such models offer a number of compelling benefits,
it has been difficult for these to compete with modern classification
algorithms in performance and even in feature extraction.
In this work, we propose a novel approach explicitly designed to address a
number of subtle yet important issues which have stymied earlier DML
algorithms. It maintains an explicit model of the distributions of the
different classes in representation space. It then employs this knowledge to
adaptively assess similarity, and achieve local discrimination by penalizing
class distribution overlap.
We demonstrate the effectiveness of this idea on several tasks. Our approach
achieves state-of-the-art classification results on a number of fine-grained
visual recognition datasets, surpassing the standard softmax classifier and
outperforming triplet loss by a relative margin of 30-40%. In terms of
computational performance, it alleviates training inefficiencies in the
traditional triplet loss, reaching the same error in 5-30 times fewer
iterations. Beyond classification, we further validate the saliency of the
learnt representations via their attribute concentration and hierarchy recovery
properties, achieving 10-25% relative gains on the softmax classifier and
25-50% on triplet loss in these tasks.
| [
{
"version": "v1",
"created": "Wed, 18 Nov 2015 20:41:05 GMT"
},
{
"version": "v2",
"created": "Wed, 2 Mar 2016 04:52:08 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Rippel",
"Oren",
""
],
[
"Paluri",
"Manohar",
""
],
[
"Dollar",
"Piotr",
""
],
[
"Bourdev",
"Lubomir",
""
]
] | TITLE: Metric Learning with Adaptive Density Discrimination
ABSTRACT: Distance metric learning (DML) approaches learn a transformation to a
representation space where distance is in correspondence with a predefined
notion of similarity. While such models offer a number of compelling benefits,
it has been difficult for these to compete with modern classification
algorithms in performance and even in feature extraction.
In this work, we propose a novel approach explicitly designed to address a
number of subtle yet important issues which have stymied earlier DML
algorithms. It maintains an explicit model of the distributions of the
different classes in representation space. It then employs this knowledge to
adaptively assess similarity, and achieve local discrimination by penalizing
class distribution overlap.
We demonstrate the effectiveness of this idea on several tasks. Our approach
achieves state-of-the-art classification results on a number of fine-grained
visual recognition datasets, surpassing the standard softmax classifier and
outperforming triplet loss by a relative margin of 30-40%. In terms of
computational performance, it alleviates training inefficiencies in the
traditional triplet loss, reaching the same error in 5-30 times fewer
iterations. Beyond classification, we further validate the saliency of the
learnt representations via their attribute concentration and hierarchy recovery
properties, achieving 10-25% relative gains on the softmax classifier and
25-50% on triplet loss in these tasks.
| no_new_dataset | 0.941654 |
1511.06238 | Miriam Cha | Miriam Cha, Youngjune Gwon, H.T. Kung | Multimodal sparse representation learning and applications | null | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Unsupervised methods have proven effective for discriminative tasks in a
single-modality scenario. In this paper, we present a multimodal framework for
learning sparse representations that can capture semantic correlation between
modalities. The framework can model relationships at a higher level by forcing
the shared sparse representation. In particular, we propose the use of joint
dictionary learning technique for sparse coding and formulate the joint
representation for concision, cross-modal representations (in case of a missing
modality), and union of the cross-modal representations. Given the accelerated
growth of multimodal data posted on the Web such as YouTube, Wikipedia, and
Twitter, learning good multimodal features is becoming increasingly important.
We show that the shared representations enabled by our framework substantially
improve the classification performance under both unimodal and multimodal
settings. We further show how deep architectures built on the proposed
framework are effective for the case of highly nonlinear correlations between
modalities. The effectiveness of our approach is demonstrated experimentally in
image denoising, multimedia event detection and retrieval on the TRECVID
dataset (audio-video), category classification on the Wikipedia dataset
(image-text), and sentiment classification on PhotoTweet (image-text).
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 16:26:24 GMT"
},
{
"version": "v2",
"created": "Sun, 24 Jan 2016 23:18:09 GMT"
},
{
"version": "v3",
"created": "Wed, 2 Mar 2016 19:22:48 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Cha",
"Miriam",
""
],
[
"Gwon",
"Youngjune",
""
],
[
"Kung",
"H. T.",
""
]
] | TITLE: Multimodal sparse representation learning and applications
ABSTRACT: Unsupervised methods have proven effective for discriminative tasks in a
single-modality scenario. In this paper, we present a multimodal framework for
learning sparse representations that can capture semantic correlation between
modalities. The framework can model relationships at a higher level by forcing
the shared sparse representation. In particular, we propose the use of joint
dictionary learning technique for sparse coding and formulate the joint
representation for concision, cross-modal representations (in case of a missing
modality), and union of the cross-modal representations. Given the accelerated
growth of multimodal data posted on the Web such as YouTube, Wikipedia, and
Twitter, learning good multimodal features is becoming increasingly important.
We show that the shared representations enabled by our framework substantially
improve the classification performance under both unimodal and multimodal
settings. We further show how deep architectures built on the proposed
framework are effective for the case of highly nonlinear correlations between
modalities. The effectiveness of our approach is demonstrated experimentally in
image denoising, multimedia event detection and retrieval on the TRECVID
dataset (audio-video), category classification on the Wikipedia dataset
(image-text), and sentiment classification on PhotoTweet (image-text).
| no_new_dataset | 0.946001 |
1511.06385 | Chunchuan Lv Mr. | Chunchuan Lyu, Kaizhu Huang, Hai-Ning Liang | A Unified Gradient Regularization Family for Adversarial Examples | The paper has been presented at ICDM 2015 | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adversarial examples are augmented data points generated by imperceptible
perturbation of input samples. They have recently drawn much attention with the
machine learning and data mining community. Being difficult to distinguish from
real examples, such adversarial examples could change the prediction of many of
the best learning models including the state-of-the-art deep learning models.
Recent attempts have been made to build robust models that take into account
adversarial examples. However, these methods can either lead to performance
drops or lack mathematical motivations. In this paper, we propose a unified
framework to build robust machine learning models against adversarial examples.
More specifically, using the unified framework, we develop a family of gradient
regularization methods that effectively penalize the gradient of loss function
w.r.t. inputs. Our proposed framework is appealing in that it offers a unified
view to deal with adversarial examples. It incorporates another
recently-proposed perturbation based approach as a special case. In addition,
we present some visual effects that reveals semantic meaning in those
perturbations, and thus support our regularization method and provide another
explanation for generalizability of adversarial examples. By applying this
technique to Maxout networks, we conduct a series of experiments and achieve
encouraging results on two benchmark datasets. In particular,we attain the best
accuracy on MNIST data (without data augmentation) and competitive performance
on CIFAR-10 data.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 21:14:43 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Lyu",
"Chunchuan",
""
],
[
"Huang",
"Kaizhu",
""
],
[
"Liang",
"Hai-Ning",
""
]
] | TITLE: A Unified Gradient Regularization Family for Adversarial Examples
ABSTRACT: Adversarial examples are augmented data points generated by imperceptible
perturbation of input samples. They have recently drawn much attention with the
machine learning and data mining community. Being difficult to distinguish from
real examples, such adversarial examples could change the prediction of many of
the best learning models including the state-of-the-art deep learning models.
Recent attempts have been made to build robust models that take into account
adversarial examples. However, these methods can either lead to performance
drops or lack mathematical motivations. In this paper, we propose a unified
framework to build robust machine learning models against adversarial examples.
More specifically, using the unified framework, we develop a family of gradient
regularization methods that effectively penalize the gradient of loss function
w.r.t. inputs. Our proposed framework is appealing in that it offers a unified
view to deal with adversarial examples. It incorporates another
recently-proposed perturbation based approach as a special case. In addition,
we present some visual effects that reveals semantic meaning in those
perturbations, and thus support our regularization method and provide another
explanation for generalizability of adversarial examples. By applying this
technique to Maxout networks, we conduct a series of experiments and achieve
encouraging results on two benchmark datasets. In particular,we attain the best
accuracy on MNIST data (without data augmentation) and competitive performance
on CIFAR-10 data.
| no_new_dataset | 0.944995 |
1603.00788 | Alp Kucukelbir | Alp Kucukelbir, Dustin Tran, Rajesh Ranganath, Andrew Gelman, David M.
Blei | Automatic Differentiation Variational Inference | null | null | null | null | stat.ML cs.AI cs.LG stat.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Probabilistic modeling is iterative. A scientist posits a simple model, fits
it to her data, refines it according to her analysis, and repeats. However,
fitting complex models to large data is a bottleneck in this process. Deriving
algorithms for new models can be both mathematically and computationally
challenging, which makes it difficult to efficiently cycle through the steps.
To this end, we develop automatic differentiation variational inference (ADVI).
Using our method, the scientist only provides a probabilistic model and a
dataset, nothing else. ADVI automatically derives an efficient variational
inference algorithm, freeing the scientist to refine and explore many models.
ADVI supports a broad class of models-no conjugacy assumptions are required. We
study ADVI across ten different models and apply it to a dataset with millions
of observations. ADVI is integrated into Stan, a probabilistic programming
system; it is available for immediate use.
| [
{
"version": "v1",
"created": "Wed, 2 Mar 2016 16:43:15 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Kucukelbir",
"Alp",
""
],
[
"Tran",
"Dustin",
""
],
[
"Ranganath",
"Rajesh",
""
],
[
"Gelman",
"Andrew",
""
],
[
"Blei",
"David M.",
""
]
] | TITLE: Automatic Differentiation Variational Inference
ABSTRACT: Probabilistic modeling is iterative. A scientist posits a simple model, fits
it to her data, refines it according to her analysis, and repeats. However,
fitting complex models to large data is a bottleneck in this process. Deriving
algorithms for new models can be both mathematically and computationally
challenging, which makes it difficult to efficiently cycle through the steps.
To this end, we develop automatic differentiation variational inference (ADVI).
Using our method, the scientist only provides a probabilistic model and a
dataset, nothing else. ADVI automatically derives an efficient variational
inference algorithm, freeing the scientist to refine and explore many models.
ADVI supports a broad class of models-no conjugacy assumptions are required. We
study ADVI across ten different models and apply it to a dataset with millions
of observations. ADVI is integrated into Stan, a probabilistic programming
system; it is available for immediate use.
| no_new_dataset | 0.948489 |
1603.00845 | Xavier Gir\'o-i-Nieto | Junting Pan, Kevin McGuinness, Elisa Sayrol, Noel O'Connor and Xavier
Giro-i-Nieto | Shallow and Deep Convolutional Networks for Saliency Prediction | Preprint of the paper accepted at 2016 IEEE Conference on Computer
Vision and Pattern Recognition (CVPR). Source code and models available at
https://github.com/imatge-upc/saliency-2016-cvpr. Junting Pan and Kevin
McGuinness contributed equally to this work | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The prediction of salient areas in images has been traditionally addressed
with hand-crafted features based on neuroscience principles. This paper,
however, addresses the problem with a completely data-driven approach by
training a convolutional neural network (convnet). The learning process is
formulated as a minimization of a loss function that measures the Euclidean
distance of the predicted saliency map with the provided ground truth. The
recent publication of large datasets of saliency prediction has provided enough
data to train end-to-end architectures that are both fast and accurate. Two
designs are proposed: a shallow convnet trained from scratch, and a another
deeper solution whose first three layers are adapted from another network
trained for classification. To the authors knowledge, these are the first
end-to-end CNNs trained and tested for the purpose of saliency prediction.
| [
{
"version": "v1",
"created": "Wed, 2 Mar 2016 19:54:02 GMT"
}
] | 2016-03-03T00:00:00 | [
[
"Pan",
"Junting",
""
],
[
"McGuinness",
"Kevin",
""
],
[
"Sayrol",
"Elisa",
""
],
[
"O'Connor",
"Noel",
""
],
[
"Giro-i-Nieto",
"Xavier",
""
]
] | TITLE: Shallow and Deep Convolutional Networks for Saliency Prediction
ABSTRACT: The prediction of salient areas in images has been traditionally addressed
with hand-crafted features based on neuroscience principles. This paper,
however, addresses the problem with a completely data-driven approach by
training a convolutional neural network (convnet). The learning process is
formulated as a minimization of a loss function that measures the Euclidean
distance of the predicted saliency map with the provided ground truth. The
recent publication of large datasets of saliency prediction has provided enough
data to train end-to-end architectures that are both fast and accurate. Two
designs are proposed: a shallow convnet trained from scratch, and a another
deeper solution whose first three layers are adapted from another network
trained for classification. To the authors knowledge, these are the first
end-to-end CNNs trained and tested for the purpose of saliency prediction.
| no_new_dataset | 0.950595 |
1503.00848 | Jordi Pont-Tuset | Jordi Pont-Tuset, Pablo Arbelaez, Jonathan T. Barron, Ferran Marques,
Jitendra Malik | Multiscale Combinatorial Grouping for Image Segmentation and Object
Proposal Generation | null | null | 10.1109/TPAMI.2016.2537320 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a unified approach for bottom-up hierarchical image segmentation
and object proposal generation for recognition, called Multiscale Combinatorial
Grouping (MCG). For this purpose, we first develop a fast normalized cuts
algorithm. We then propose a high-performance hierarchical segmenter that makes
effective use of multiscale information. Finally, we propose a grouping
strategy that combines our multiscale regions into highly-accurate object
proposals by exploring efficiently their combinatorial space. We also present
Single-scale Combinatorial Grouping (SCG), a faster version of MCG that
produces competitive proposals in under five second per image. We conduct an
extensive and comprehensive empirical validation on the BSDS500, SegVOC12, SBD,
and COCO datasets, showing that MCG produces state-of-the-art contours,
hierarchical regions, and object proposals.
| [
{
"version": "v1",
"created": "Tue, 3 Mar 2015 07:58:22 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Mar 2015 12:00:33 GMT"
},
{
"version": "v3",
"created": "Mon, 29 Jun 2015 17:57:06 GMT"
},
{
"version": "v4",
"created": "Tue, 1 Mar 2016 09:00:09 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Pont-Tuset",
"Jordi",
""
],
[
"Arbelaez",
"Pablo",
""
],
[
"Barron",
"Jonathan T.",
""
],
[
"Marques",
"Ferran",
""
],
[
"Malik",
"Jitendra",
""
]
] | TITLE: Multiscale Combinatorial Grouping for Image Segmentation and Object
Proposal Generation
ABSTRACT: We propose a unified approach for bottom-up hierarchical image segmentation
and object proposal generation for recognition, called Multiscale Combinatorial
Grouping (MCG). For this purpose, we first develop a fast normalized cuts
algorithm. We then propose a high-performance hierarchical segmenter that makes
effective use of multiscale information. Finally, we propose a grouping
strategy that combines our multiscale regions into highly-accurate object
proposals by exploring efficiently their combinatorial space. We also present
Single-scale Combinatorial Grouping (SCG), a faster version of MCG that
produces competitive proposals in under five second per image. We conduct an
extensive and comprehensive empirical validation on the BSDS500, SegVOC12, SBD,
and COCO datasets, showing that MCG produces state-of-the-art contours,
hierarchical regions, and object proposals.
| no_new_dataset | 0.955775 |
1507.04760 | Lex Fridman | Lex Fridman, Philipp Langhans, Joonbum Lee, Bryan Reimer | Driver Gaze Region Estimation Without Using Eye Movement | Accepted for Publication in IEEE Intelligent Systems | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated estimation of the allocation of a driver's visual attention may be
a critical component of future Advanced Driver Assistance Systems. In theory,
vision-based tracking of the eye can provide a good estimate of gaze location.
In practice, eye tracking from video is challenging because of sunglasses,
eyeglass reflections, lighting conditions, occlusions, motion blur, and other
factors. Estimation of head pose, on the other hand, is robust to many of these
effects, but cannot provide as fine-grained of a resolution in localizing the
gaze. However, for the purpose of keeping the driver safe, it is sufficient to
partition gaze into regions. In this effort, we propose a system that extracts
facial features and classifies their spatial configuration into six regions in
real-time. Our proposed method achieves an average accuracy of 91.4% at an
average decision rate of 11 Hz on a dataset of 50 drivers from an on-road
study.
| [
{
"version": "v1",
"created": "Thu, 16 Jul 2015 20:16:20 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Mar 2016 17:21:25 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Fridman",
"Lex",
""
],
[
"Langhans",
"Philipp",
""
],
[
"Lee",
"Joonbum",
""
],
[
"Reimer",
"Bryan",
""
]
] | TITLE: Driver Gaze Region Estimation Without Using Eye Movement
ABSTRACT: Automated estimation of the allocation of a driver's visual attention may be
a critical component of future Advanced Driver Assistance Systems. In theory,
vision-based tracking of the eye can provide a good estimate of gaze location.
In practice, eye tracking from video is challenging because of sunglasses,
eyeglass reflections, lighting conditions, occlusions, motion blur, and other
factors. Estimation of head pose, on the other hand, is robust to many of these
effects, but cannot provide as fine-grained of a resolution in localizing the
gaze. However, for the purpose of keeping the driver safe, it is sufficient to
partition gaze into regions. In this effort, we propose a system that extracts
facial features and classifies their spatial configuration into six regions in
real-time. Our proposed method achieves an average accuracy of 91.4% at an
average decision rate of 11 Hz on a dataset of 50 drivers from an on-road
study.
| no_new_dataset | 0.940134 |
1509.06664 | Tim Rockt\"aschel | Tim Rockt\"aschel, Edward Grefenstette, Karl Moritz Hermann,
Tom\'a\v{s} Ko\v{c}isk\'y, Phil Blunsom | Reasoning about Entailment with Neural Attention | ICLR 2016 camera-ready, 9 pages, 10 figures (incl. subfigures) | null | null | null | cs.CL cs.AI cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While most approaches to automatically recognizing entailment relations have
used classifiers employing hand engineered features derived from complex
natural language processing pipelines, in practice their performance has been
only slightly better than bag-of-word pair classifiers using only lexical
similarity. The only attempt so far to build an end-to-end differentiable
neural network for entailment failed to outperform such a simple similarity
classifier. In this paper, we propose a neural model that reads two sentences
to determine entailment using long short-term memory units. We extend this
model with a word-by-word neural attention mechanism that encourages reasoning
over entailments of pairs of words and phrases. Furthermore, we present a
qualitative analysis of attention weights produced by this model, demonstrating
such reasoning capabilities. On a large entailment dataset this model
outperforms the previous best neural model and a classifier with engineered
features by a substantial margin. It is the first generic end-to-end
differentiable system that achieves state-of-the-art accuracy on a textual
entailment dataset.
| [
{
"version": "v1",
"created": "Tue, 22 Sep 2015 16:08:24 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Nov 2015 22:12:52 GMT"
},
{
"version": "v3",
"created": "Mon, 18 Jan 2016 17:28:30 GMT"
},
{
"version": "v4",
"created": "Tue, 1 Mar 2016 10:32:06 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Rocktäschel",
"Tim",
""
],
[
"Grefenstette",
"Edward",
""
],
[
"Hermann",
"Karl Moritz",
""
],
[
"Kočiský",
"Tomáš",
""
],
[
"Blunsom",
"Phil",
""
]
] | TITLE: Reasoning about Entailment with Neural Attention
ABSTRACT: While most approaches to automatically recognizing entailment relations have
used classifiers employing hand engineered features derived from complex
natural language processing pipelines, in practice their performance has been
only slightly better than bag-of-word pair classifiers using only lexical
similarity. The only attempt so far to build an end-to-end differentiable
neural network for entailment failed to outperform such a simple similarity
classifier. In this paper, we propose a neural model that reads two sentences
to determine entailment using long short-term memory units. We extend this
model with a word-by-word neural attention mechanism that encourages reasoning
over entailments of pairs of words and phrases. Furthermore, we present a
qualitative analysis of attention weights produced by this model, demonstrating
such reasoning capabilities. On a large entailment dataset this model
outperforms the previous best neural model and a classifier with engineered
features by a substantial margin. It is the first generic end-to-end
differentiable system that achieves state-of-the-art accuracy on a textual
entailment dataset.
| no_new_dataset | 0.946349 |
1511.06432 | Nicolas Ballas | Nicolas Ballas, Li Yao, Chris Pal, Aaron Courville | Delving Deeper into Convolutional Networks for Learning Video
Representations | ICLR 2016 | null | null | null | cs.CV cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an approach to learn spatio-temporal features in videos from
intermediate visual representations we call "percepts" using
Gated-Recurrent-Unit Recurrent Networks (GRUs).Our method relies on percepts
that are extracted from all level of a deep convolutional network trained on
the large ImageNet dataset. While high-level percepts contain highly
discriminative information, they tend to have a low-spatial resolution.
Low-level percepts, on the other hand, preserve a higher spatial resolution
from which we can model finer motion patterns. Using low-level percepts can
leads to high-dimensionality video representations. To mitigate this effect and
control the model number of parameters, we introduce a variant of the GRU model
that leverages the convolution operations to enforce sparse connectivity of the
model units and share parameters across the input spatial locations.
We empirically validate our approach on both Human Action Recognition and
Video Captioning tasks. In particular, we achieve results equivalent to
state-of-art on the YouTube2Text dataset using a simpler text-decoder model and
without extra 3D CNN features.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 22:46:13 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Nov 2015 02:46:54 GMT"
},
{
"version": "v3",
"created": "Thu, 7 Jan 2016 19:43:19 GMT"
},
{
"version": "v4",
"created": "Tue, 1 Mar 2016 18:54:11 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Ballas",
"Nicolas",
""
],
[
"Yao",
"Li",
""
],
[
"Pal",
"Chris",
""
],
[
"Courville",
"Aaron",
""
]
] | TITLE: Delving Deeper into Convolutional Networks for Learning Video
Representations
ABSTRACT: We propose an approach to learn spatio-temporal features in videos from
intermediate visual representations we call "percepts" using
Gated-Recurrent-Unit Recurrent Networks (GRUs).Our method relies on percepts
that are extracted from all level of a deep convolutional network trained on
the large ImageNet dataset. While high-level percepts contain highly
discriminative information, they tend to have a low-spatial resolution.
Low-level percepts, on the other hand, preserve a higher spatial resolution
from which we can model finer motion patterns. Using low-level percepts can
leads to high-dimensionality video representations. To mitigate this effect and
control the model number of parameters, we introduce a variant of the GRU model
that leverages the convolution operations to enforce sparse connectivity of the
model units and share parameters across the input spatial locations.
We empirically validate our approach on both Human Action Recognition and
Video Captioning tasks. In particular, we achieve results equivalent to
state-of-art on the YouTube2Text dataset using a simpler text-decoder model and
without extra 3D CNN features.
| no_new_dataset | 0.9462 |
1511.06455 | Zhenwen Dai | Zhenwen Dai, Andreas Damianou, Javier Gonz\'alez, Neil Lawrence | Variational Auto-encoded Deep Gaussian Processes | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a scalable deep non-parametric generative model by augmenting deep
Gaussian processes with a recognition model. Inference is performed in a novel
scalable variational framework where the variational posterior distributions
are reparametrized through a multilayer perceptron. The key aspect of this
reformulation is that it prevents the proliferation of variational parameters
which otherwise grow linearly in proportion to the sample size. We derive a new
formulation of the variational lower bound that allows us to distribute most of
the computation in a way that enables to handle datasets of the size of
mainstream deep learning tasks. We show the efficacy of the method on a variety
of challenges including deep unsupervised learning and deep Bayesian
optimization.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 23:47:34 GMT"
},
{
"version": "v2",
"created": "Mon, 29 Feb 2016 21:34:58 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Dai",
"Zhenwen",
""
],
[
"Damianou",
"Andreas",
""
],
[
"González",
"Javier",
""
],
[
"Lawrence",
"Neil",
""
]
] | TITLE: Variational Auto-encoded Deep Gaussian Processes
ABSTRACT: We develop a scalable deep non-parametric generative model by augmenting deep
Gaussian processes with a recognition model. Inference is performed in a novel
scalable variational framework where the variational posterior distributions
are reparametrized through a multilayer perceptron. The key aspect of this
reformulation is that it prevents the proliferation of variational parameters
which otherwise grow linearly in proportion to the sample size. We derive a new
formulation of the variational lower bound that allows us to distribute most of
the computation in a way that enables to handle datasets of the size of
mainstream deep learning tasks. We show the efficacy of the method on a variety
of challenges including deep unsupervised learning and deep Bayesian
optimization.
| no_new_dataset | 0.944638 |
1603.00016 | Gerald Eigen | Gerald Eigen, Are Tr{\ae}et, Justas Zalieckas, Jaroslav Cvach, Jiri
Kvasnicka, Ivo Polak | SiPM Gain Stabilization Studies for Adaptive Power Supply | 14 pages, 41 figures, Talk presented at the International Workshop on
Future Linear Colliders (LCWS15), Whistler, Canada, 2-6 November 2015 | null | null | null | physics.ins-det hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present herein gain stabilization studies of SiPMs using a climate chamber
at CERN. We present results for four detectors not tested before, three from
Hamamatsu and one from KETEK. Two of the Hamamatsu SiPMs are novel sensors with
trenches that reduce cross talk. We use an improved readout system with a
digital oscilloscope controlled with a dedicated LabView program. We improved
and automized the analysis to deal with large datasets. We have measured the
gain-versus-bias-voltage dependence at fixed temperature and
gain-versus-temperature dependence at fixed bias voltage to determine the bias
voltage dependence on temperature $V(T)$ for stable gain. We show that the gain
remains stable to better than $\pm 0.5\%$ in the $20^\circ \rm C - 30^\circ C$
temperature range if the bias voltage is properly adjusted with temperature.
| [
{
"version": "v1",
"created": "Mon, 29 Feb 2016 19:44:24 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Eigen",
"Gerald",
""
],
[
"Træet",
"Are",
""
],
[
"Zalieckas",
"Justas",
""
],
[
"Cvach",
"Jaroslav",
""
],
[
"Kvasnicka",
"Jiri",
""
],
[
"Polak",
"Ivo",
""
]
] | TITLE: SiPM Gain Stabilization Studies for Adaptive Power Supply
ABSTRACT: We present herein gain stabilization studies of SiPMs using a climate chamber
at CERN. We present results for four detectors not tested before, three from
Hamamatsu and one from KETEK. Two of the Hamamatsu SiPMs are novel sensors with
trenches that reduce cross talk. We use an improved readout system with a
digital oscilloscope controlled with a dedicated LabView program. We improved
and automized the analysis to deal with large datasets. We have measured the
gain-versus-bias-voltage dependence at fixed temperature and
gain-versus-temperature dependence at fixed bias voltage to determine the bias
voltage dependence on temperature $V(T)$ for stable gain. We show that the gain
remains stable to better than $\pm 0.5\%$ in the $20^\circ \rm C - 30^\circ C$
temperature range if the bias voltage is properly adjusted with temperature.
| no_new_dataset | 0.938181 |
1603.00128 | Yanwei Pang | Xiaoheng Jiang, Yanwei Pang, Manli Sun, and Xuelong Li | Cascaded Subpatch Networks for Effective CNNs | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Conventional Convolutional Neural Networks (CNNs) use either a linear or
non-linear filter to extract features from an image patch (region) of spatial
size $ H\times W $ (Typically, $ H $ is small and is equal to $ W$, e.g., $ H $
is 5 or 7). Generally, the size of the filter is equal to the size $ H\times W
$ of the input patch. We argue that the representation ability of equal-size
strategy is not strong enough. To overcome the drawback, we propose to use
subpatch filter whose spatial size $ h\times w $ is smaller than $ H\times W $.
The proposed subpatch filter consists of two subsequent filters. The first one
is a linear filter of spatial size $ h\times w $ and is aimed at extracting
features from spatial domain. The second one is of spatial size $ 1\times 1 $
and is used for strengthening the connection between different input feature
channels and for reducing the number of parameters. The subpatch filter
convolves with the input patch and the resulting network is called a subpatch
network. Taking the output of one subpatch network as input, we further repeat
constructing subpatch networks until the output contains only one neuron in
spatial domain. These subpatch networks form a new network called Cascaded
Subpatch Network (CSNet). The feature layer generated by CSNet is called csconv
layer. For the whole input image, we construct a deep neural network by
stacking a sequence of csconv layers. Experimental results on four benchmark
datasets demonstrate the effectiveness and compactness of the proposed CSNet.
For example, our CSNet reaches a test error of $ 5.68\% $ on the CIFAR10
dataset without model averaging. To the best of our knowledge, this is the best
result ever obtained on the CIFAR10 dataset.
| [
{
"version": "v1",
"created": "Tue, 1 Mar 2016 03:44:49 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Jiang",
"Xiaoheng",
""
],
[
"Pang",
"Yanwei",
""
],
[
"Sun",
"Manli",
""
],
[
"Li",
"Xuelong",
""
]
] | TITLE: Cascaded Subpatch Networks for Effective CNNs
ABSTRACT: Conventional Convolutional Neural Networks (CNNs) use either a linear or
non-linear filter to extract features from an image patch (region) of spatial
size $ H\times W $ (Typically, $ H $ is small and is equal to $ W$, e.g., $ H $
is 5 or 7). Generally, the size of the filter is equal to the size $ H\times W
$ of the input patch. We argue that the representation ability of equal-size
strategy is not strong enough. To overcome the drawback, we propose to use
subpatch filter whose spatial size $ h\times w $ is smaller than $ H\times W $.
The proposed subpatch filter consists of two subsequent filters. The first one
is a linear filter of spatial size $ h\times w $ and is aimed at extracting
features from spatial domain. The second one is of spatial size $ 1\times 1 $
and is used for strengthening the connection between different input feature
channels and for reducing the number of parameters. The subpatch filter
convolves with the input patch and the resulting network is called a subpatch
network. Taking the output of one subpatch network as input, we further repeat
constructing subpatch networks until the output contains only one neuron in
spatial domain. These subpatch networks form a new network called Cascaded
Subpatch Network (CSNet). The feature layer generated by CSNet is called csconv
layer. For the whole input image, we construct a deep neural network by
stacking a sequence of csconv layers. Experimental results on four benchmark
datasets demonstrate the effectiveness and compactness of the proposed CSNet.
For example, our CSNet reaches a test error of $ 5.68\% $ on the CIFAR10
dataset without model averaging. To the best of our knowledge, this is the best
result ever obtained on the CIFAR10 dataset.
| no_new_dataset | 0.951953 |
1603.00150 | Dylan Campbell | Dylan Campbell and Lars Petersson | GOGMA: Globally-Optimal Gaussian Mixture Alignment | Manuscript in press 2016 IEEE Conference on Computer Vision and
Pattern Recognition | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gaussian mixture alignment is a family of approaches that are frequently used
for robustly solving the point-set registration problem. However, since they
use local optimisation, they are susceptible to local minima and can only
guarantee local optimality. Consequently, their accuracy is strongly dependent
on the quality of the initialisation. This paper presents the first
globally-optimal solution to the 3D rigid Gaussian mixture alignment problem
under the L2 distance between mixtures. The algorithm, named GOGMA, employs a
branch-and-bound approach to search the space of 3D rigid motions SE(3),
guaranteeing global optimality regardless of the initialisation. The geometry
of SE(3) was used to find novel upper and lower bounds for the objective
function and local optimisation was integrated into the scheme to accelerate
convergence without voiding the optimality guarantee. The evaluation
empirically supported the optimality proof and showed that the method performed
much more robustly on two challenging datasets than an existing
globally-optimal registration solution.
| [
{
"version": "v1",
"created": "Tue, 1 Mar 2016 05:38:50 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Campbell",
"Dylan",
""
],
[
"Petersson",
"Lars",
""
]
] | TITLE: GOGMA: Globally-Optimal Gaussian Mixture Alignment
ABSTRACT: Gaussian mixture alignment is a family of approaches that are frequently used
for robustly solving the point-set registration problem. However, since they
use local optimisation, they are susceptible to local minima and can only
guarantee local optimality. Consequently, their accuracy is strongly dependent
on the quality of the initialisation. This paper presents the first
globally-optimal solution to the 3D rigid Gaussian mixture alignment problem
under the L2 distance between mixtures. The algorithm, named GOGMA, employs a
branch-and-bound approach to search the space of 3D rigid motions SE(3),
guaranteeing global optimality regardless of the initialisation. The geometry
of SE(3) was used to find novel upper and lower bounds for the objective
function and local optimisation was integrated into the scheme to accelerate
convergence without voiding the optimality guarantee. The evaluation
empirically supported the optimality proof and showed that the method performed
much more robustly on two challenging datasets than an existing
globally-optimal registration solution.
| no_new_dataset | 0.950319 |
1603.00431 | Amirali Sanatinia | Amirali Sanatinia, Guevara Noubir | On GitHub's Programming Languages | null | null | null | null | cs.PL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | GitHub is the most widely used social, distributed version control system. It
has around 10 million registered users and hosts over 16 million public
repositories. Its user base is also very active as GitHub ranks in the top 100
Alexa most popular websites. In this study, we collect GitHub's state in its
entirety. Doing so, allows us to study new aspects of the ecosystem. Although
GitHub is the home to millions of users and repositories, the analysis of
users' activity time-series reveals that only around 10% of them can be
considered active. The collected dataset allows us to investigate the
popularity of programming languages and existence of pattens in the relations
between users, repositories, and programming languages.
By, applying a k-means clustering method to the users-repositories commits
matrix, we find that two clear clusters of programming languages separate from
the remaining. One cluster forms for "web programming" languages (Java Script,
Ruby, PHP, CSS), and a second for "system oriented programming" languages (C,
C++, Python). Further classification, allow us to build a phylogenetic tree of
the use of programming languages in GitHub. Additionally, we study the main and
the auxiliary programming languages of the top 1000 repositories in more
detail. We provide a ranking of these auxiliary programming languages using
various metrics, such as percentage of lines of code, and PageRank.
| [
{
"version": "v1",
"created": "Tue, 1 Mar 2016 20:03:44 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Sanatinia",
"Amirali",
""
],
[
"Noubir",
"Guevara",
""
]
] | TITLE: On GitHub's Programming Languages
ABSTRACT: GitHub is the most widely used social, distributed version control system. It
has around 10 million registered users and hosts over 16 million public
repositories. Its user base is also very active as GitHub ranks in the top 100
Alexa most popular websites. In this study, we collect GitHub's state in its
entirety. Doing so, allows us to study new aspects of the ecosystem. Although
GitHub is the home to millions of users and repositories, the analysis of
users' activity time-series reveals that only around 10% of them can be
considered active. The collected dataset allows us to investigate the
popularity of programming languages and existence of pattens in the relations
between users, repositories, and programming languages.
By, applying a k-means clustering method to the users-repositories commits
matrix, we find that two clear clusters of programming languages separate from
the remaining. One cluster forms for "web programming" languages (Java Script,
Ruby, PHP, CSS), and a second for "system oriented programming" languages (C,
C++, Python). Further classification, allow us to build a phylogenetic tree of
the use of programming languages in GitHub. Additionally, we study the main and
the auxiliary programming languages of the top 1000 repositories in more
detail. We provide a ranking of these auxiliary programming languages using
various metrics, such as percentage of lines of code, and PageRank.
| new_dataset | 0.92912 |
1603.00438 | Team Lear | Mattis Paulin (LEAR), Julien Mairal (LEAR), Matthijs Douze (LEAR),
Zaid Harchaoui (NYU), Florent Perronnin, Cordelia Schmid (LEAR) | Convolutional Patch Representations for Image Retrieval: an Unsupervised
Approach | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Convolutional neural networks (CNNs) have recently received a lot of
attention due to their ability to model local stationary structures in natural
images in a multi-scale fashion, when learning all model parameters with
supervision. While excellent performance was achieved for image classification
when large amounts of labeled visual data are available, their success for
un-supervised tasks such as image retrieval has been moderate so far. Our paper
focuses on this latter setting and explores several methods for learning patch
descriptors without supervision with application to matching and instance-level
retrieval. To that effect, we propose a new family of convolutional descriptors
for patch representation , based on the recently introduced convolutional
kernel networks. We show that our descriptor, named Patch-CKN, performs better
than SIFT as well as other convolutional networks learned by artificially
introducing supervision and is significantly faster to train. To demonstrate
its effectiveness, we perform an extensive evaluation on standard benchmarks
for patch and image retrieval where we obtain state-of-the-art results. We also
introduce a new dataset called RomePatches, which allows to simultaneously
study descriptor performance for patch and image retrieval.
| [
{
"version": "v1",
"created": "Tue, 1 Mar 2016 20:13:07 GMT"
}
] | 2016-03-02T00:00:00 | [
[
"Paulin",
"Mattis",
"",
"LEAR"
],
[
"Mairal",
"Julien",
"",
"LEAR"
],
[
"Douze",
"Matthijs",
"",
"LEAR"
],
[
"Harchaoui",
"Zaid",
"",
"NYU"
],
[
"Perronnin",
"Florent",
"",
"LEAR"
],
[
"Schmid",
"Cordelia",
"",
"LEAR"
]
] | TITLE: Convolutional Patch Representations for Image Retrieval: an Unsupervised
Approach
ABSTRACT: Convolutional neural networks (CNNs) have recently received a lot of
attention due to their ability to model local stationary structures in natural
images in a multi-scale fashion, when learning all model parameters with
supervision. While excellent performance was achieved for image classification
when large amounts of labeled visual data are available, their success for
un-supervised tasks such as image retrieval has been moderate so far. Our paper
focuses on this latter setting and explores several methods for learning patch
descriptors without supervision with application to matching and instance-level
retrieval. To that effect, we propose a new family of convolutional descriptors
for patch representation , based on the recently introduced convolutional
kernel networks. We show that our descriptor, named Patch-CKN, performs better
than SIFT as well as other convolutional networks learned by artificially
introducing supervision and is significantly faster to train. To demonstrate
its effectiveness, we perform an extensive evaluation on standard benchmarks
for patch and image retrieval where we obtain state-of-the-art results. We also
introduce a new dataset called RomePatches, which allows to simultaneously
study descriptor performance for patch and image retrieval.
| new_dataset | 0.954393 |
1407.3422 | Igor Melnyk | Igor Melnyk and Arindam Banerjee | A Spectral Algorithm for Inference in Hidden Semi-Markov Models | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hidden semi-Markov models (HSMMs) are latent variable models which allow
latent state persistence and can be viewed as a generalization of the popular
hidden Markov models (HMMs). In this paper, we introduce a novel spectral
algorithm to perform inference in HSMMs. Unlike expectation maximization (EM),
our approach correctly estimates the probability of given observation sequence
based on a set of training sequences. Our approach is based on estimating
moments from the sample, whose number of dimensions depends only
logarithmically on the maximum length of the hidden state persistence.
Moreover, the algorithm requires only a few matrix inversions and is therefore
computationally efficient. Empirical evaluations on synthetic and real data
demonstrate the advantage of the algorithm over EM in terms of speed and
accuracy, especially for large datasets.
| [
{
"version": "v1",
"created": "Sat, 12 Jul 2014 23:57:07 GMT"
},
{
"version": "v2",
"created": "Sun, 21 Feb 2016 19:29:03 GMT"
},
{
"version": "v3",
"created": "Mon, 29 Feb 2016 00:29:23 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Melnyk",
"Igor",
""
],
[
"Banerjee",
"Arindam",
""
]
] | TITLE: A Spectral Algorithm for Inference in Hidden Semi-Markov Models
ABSTRACT: Hidden semi-Markov models (HSMMs) are latent variable models which allow
latent state persistence and can be viewed as a generalization of the popular
hidden Markov models (HMMs). In this paper, we introduce a novel spectral
algorithm to perform inference in HSMMs. Unlike expectation maximization (EM),
our approach correctly estimates the probability of given observation sequence
based on a set of training sequences. Our approach is based on estimating
moments from the sample, whose number of dimensions depends only
logarithmically on the maximum length of the hidden state persistence.
Moreover, the algorithm requires only a few matrix inversions and is therefore
computationally efficient. Empirical evaluations on synthetic and real data
demonstrate the advantage of the algorithm over EM in terms of speed and
accuracy, especially for large datasets.
| no_new_dataset | 0.949716 |
1506.05907 | Bastien Mussard | Bastien Mussard (LCT, ICS), Peter Reinhardt (LCT), Janos Angyan (UL),
Julien Toulouse (LCT) | Spin-unrestricted random-phase approximation with range separation:
Benchmark on atomization energies and reaction barrier heights | null | Journal of Chemical Physics, American Institute of Physics, 2015,
pp.00 | 10.1063/1.4918710 | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider several spin-unrestricted random-phase approximation (RPA)
variants for calculating correlation energies, with and without range
separation, and test them on datasets of atomization energies and reaction
barrier heights. We show that range separation greatly improves the accuracy of
all RPA variants for these properties. Moreover, we show that a RPA variant
with exchange, hereafter referred to as RPAx-SO2, first proposed by Sz-abo and
Ostlund [A. Szabo and N. S. Ostlund, J. Chem. Phys. 67, 4351 (1977)] in a
spin-restricted closed-shell formalism, and extended here to a
spin-unrestricted formalism , provides on average the most accurate
range-separated RPA variant for atomization energies and reaction barrier
heights. Since this range-separated RPAx-SO2 method had already been shown to
be among the most accurate range-separated RPA variants for weak intermolecular
interactions [J. Toulouse, W. Zhu, A. Savin, G. Jansen, and J. G.
{\'A}ngy{\'a}n, J. Chem. Phys. 135, 084119 (2011)], this works confirms
range-separated RPAx-SO2 as a promising method for general chemical
applications.
| [
{
"version": "v1",
"created": "Fri, 19 Jun 2015 08:27:25 GMT"
},
{
"version": "v2",
"created": "Fri, 26 Feb 2016 14:07:17 GMT"
},
{
"version": "v3",
"created": "Mon, 29 Feb 2016 19:42:36 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Mussard",
"Bastien",
"",
"LCT, ICS"
],
[
"Reinhardt",
"Peter",
"",
"LCT"
],
[
"Angyan",
"Janos",
"",
"UL"
],
[
"Toulouse",
"Julien",
"",
"LCT"
]
] | TITLE: Spin-unrestricted random-phase approximation with range separation:
Benchmark on atomization energies and reaction barrier heights
ABSTRACT: We consider several spin-unrestricted random-phase approximation (RPA)
variants for calculating correlation energies, with and without range
separation, and test them on datasets of atomization energies and reaction
barrier heights. We show that range separation greatly improves the accuracy of
all RPA variants for these properties. Moreover, we show that a RPA variant
with exchange, hereafter referred to as RPAx-SO2, first proposed by Sz-abo and
Ostlund [A. Szabo and N. S. Ostlund, J. Chem. Phys. 67, 4351 (1977)] in a
spin-restricted closed-shell formalism, and extended here to a
spin-unrestricted formalism , provides on average the most accurate
range-separated RPA variant for atomization energies and reaction barrier
heights. Since this range-separated RPAx-SO2 method had already been shown to
be among the most accurate range-separated RPA variants for weak intermolecular
interactions [J. Toulouse, W. Zhu, A. Savin, G. Jansen, and J. G.
{\'A}ngy{\'a}n, J. Chem. Phys. 135, 084119 (2011)], this works confirms
range-separated RPAx-SO2 as a promising method for general chemical
applications.
| no_new_dataset | 0.951142 |
1511.02793 | Elman Mansimov | Elman Mansimov, Emilio Parisotto, Jimmy Lei Ba, Ruslan Salakhutdinov | Generating Images from Captions with Attention | Published as a conference paper at ICLR 2016 | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motivated by the recent progress in generative models, we introduce a model
that generates images from natural language descriptions. The proposed model
iteratively draws patches on a canvas, while attending to the relevant words in
the description. After training on Microsoft COCO, we compare our model with
several baseline generative models on image generation and retrieval tasks. We
demonstrate that our model produces higher quality samples than other
approaches and generates images with novel scene compositions corresponding to
previously unseen captions in the dataset.
| [
{
"version": "v1",
"created": "Mon, 9 Nov 2015 18:18:53 GMT"
},
{
"version": "v2",
"created": "Mon, 29 Feb 2016 17:56:29 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Mansimov",
"Elman",
""
],
[
"Parisotto",
"Emilio",
""
],
[
"Ba",
"Jimmy Lei",
""
],
[
"Salakhutdinov",
"Ruslan",
""
]
] | TITLE: Generating Images from Captions with Attention
ABSTRACT: Motivated by the recent progress in generative models, we introduce a model
that generates images from natural language descriptions. The proposed model
iteratively draws patches on a canvas, while attending to the relevant words in
the description. After training on Microsoft COCO, we compare our model with
several baseline generative models on image generation and retrieval tasks. We
demonstrate that our model produces higher quality samples than other
approaches and generates images with novel scene compositions corresponding to
previously unseen captions in the dataset.
| no_new_dataset | 0.960398 |
1511.04773 | Weiran Wang | Weiran Wang, Karen Livescu | Large-Scale Approximate Kernel Canonical Correlation Analysis | Published as a conference paper at International Conference on
Learning Representations (ICLR) 2016 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Kernel canonical correlation analysis (KCCA) is a nonlinear multi-view
representation learning technique with broad applicability in statistics and
machine learning. Although there is a closed-form solution for the KCCA
objective, it involves solving an $N\times N$ eigenvalue system where $N$ is
the training set size, making its computational requirements in both memory and
time prohibitive for large-scale problems. Various approximation techniques
have been developed for KCCA. A commonly used approach is to first transform
the original inputs to an $M$-dimensional random feature space so that inner
products in the feature space approximate kernel evaluations, and then apply
linear CCA to the transformed inputs. In many applications, however, the
dimensionality $M$ of the random feature space may need to be very large in
order to obtain a sufficiently good approximation; it then becomes challenging
to perform the linear CCA step on the resulting very high-dimensional data
matrices. We show how to use a stochastic optimization algorithm, recently
proposed for linear CCA and its neural-network extension, to further alleviate
the computation requirements of approximate KCCA. This approach allows us to
run approximate KCCA on a speech dataset with $1.4$ million training samples
and a random feature space of dimensionality $M=100000$ on a typical
workstation.
| [
{
"version": "v1",
"created": "Sun, 15 Nov 2015 22:20:02 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Nov 2015 16:31:14 GMT"
},
{
"version": "v3",
"created": "Thu, 7 Jan 2016 00:27:20 GMT"
},
{
"version": "v4",
"created": "Mon, 29 Feb 2016 16:04:46 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Wang",
"Weiran",
""
],
[
"Livescu",
"Karen",
""
]
] | TITLE: Large-Scale Approximate Kernel Canonical Correlation Analysis
ABSTRACT: Kernel canonical correlation analysis (KCCA) is a nonlinear multi-view
representation learning technique with broad applicability in statistics and
machine learning. Although there is a closed-form solution for the KCCA
objective, it involves solving an $N\times N$ eigenvalue system where $N$ is
the training set size, making its computational requirements in both memory and
time prohibitive for large-scale problems. Various approximation techniques
have been developed for KCCA. A commonly used approach is to first transform
the original inputs to an $M$-dimensional random feature space so that inner
products in the feature space approximate kernel evaluations, and then apply
linear CCA to the transformed inputs. In many applications, however, the
dimensionality $M$ of the random feature space may need to be very large in
order to obtain a sufficiently good approximation; it then becomes challenging
to perform the linear CCA step on the resulting very high-dimensional data
matrices. We show how to use a stochastic optimization algorithm, recently
proposed for linear CCA and its neural-network extension, to further alleviate
the computation requirements of approximate KCCA. This approach allows us to
run approximate KCCA on a speech dataset with $1.4$ million training samples
and a random feature space of dimensionality $M=100000$ on a typical
workstation.
| no_new_dataset | 0.941815 |
1511.05440 | Michael Mathieu | Michael Mathieu, Camille Couprie and Yann LeCun | Deep multi-scale video prediction beyond mean square error | null | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning to predict future images from a video sequence involves the
construction of an internal representation that models the image evolution
accurately, and therefore, to some degree, its content and dynamics. This is
why pixel-space video prediction may be viewed as a promising avenue for
unsupervised feature learning. In addition, while optical flow has been a very
studied problem in computer vision for a long time, future frame prediction is
rarely approached. Still, many vision applications could benefit from the
knowledge of the next frames of videos, that does not require the complexity of
tracking every pixel trajectories. In this work, we train a convolutional
network to generate future frames given an input sequence. To deal with the
inherently blurry predictions obtained from the standard Mean Squared Error
(MSE) loss function, we propose three different and complementary feature
learning strategies: a multi-scale architecture, an adversarial training
method, and an image gradient difference loss function. We compare our
predictions to different published results based on recurrent neural networks
on the UCF101 dataset
| [
{
"version": "v1",
"created": "Tue, 17 Nov 2015 15:36:32 GMT"
},
{
"version": "v2",
"created": "Thu, 19 Nov 2015 23:21:22 GMT"
},
{
"version": "v3",
"created": "Mon, 23 Nov 2015 04:58:24 GMT"
},
{
"version": "v4",
"created": "Thu, 7 Jan 2016 21:52:53 GMT"
},
{
"version": "v5",
"created": "Fri, 15 Jan 2016 02:09:16 GMT"
},
{
"version": "v6",
"created": "Fri, 26 Feb 2016 22:10:30 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Mathieu",
"Michael",
""
],
[
"Couprie",
"Camille",
""
],
[
"LeCun",
"Yann",
""
]
] | TITLE: Deep multi-scale video prediction beyond mean square error
ABSTRACT: Learning to predict future images from a video sequence involves the
construction of an internal representation that models the image evolution
accurately, and therefore, to some degree, its content and dynamics. This is
why pixel-space video prediction may be viewed as a promising avenue for
unsupervised feature learning. In addition, while optical flow has been a very
studied problem in computer vision for a long time, future frame prediction is
rarely approached. Still, many vision applications could benefit from the
knowledge of the next frames of videos, that does not require the complexity of
tracking every pixel trajectories. In this work, we train a convolutional
network to generate future frames given an input sequence. To deal with the
inherently blurry predictions obtained from the standard Mean Squared Error
(MSE) loss function, we propose three different and complementary feature
learning strategies: a multi-scale architecture, an adversarial training
method, and an image gradient difference loss function. We compare our
predictions to different published results based on recurrent neural networks
on the UCF101 dataset
| no_new_dataset | 0.949342 |
1511.06051 | Robert Nishihara | Philipp Moritz, Robert Nishihara, Ion Stoica, Michael I. Jordan | SparkNet: Training Deep Networks in Spark | 12 pages, 7 figures | null | null | null | stat.ML cs.DC cs.LG cs.NE math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Training deep networks is a time-consuming process, with networks for object
recognition often requiring multiple days to train. For this reason, leveraging
the resources of a cluster to speed up training is an important area of work.
However, widely-popular batch-processing computational frameworks like
MapReduce and Spark were not designed to support the asynchronous and
communication-intensive workloads of existing distributed deep learning
systems. We introduce SparkNet, a framework for training deep networks in
Spark. Our implementation includes a convenient interface for reading data from
Spark RDDs, a Scala interface to the Caffe deep learning framework, and a
lightweight multi-dimensional tensor library. Using a simple parallelization
scheme for stochastic gradient descent, SparkNet scales well with the cluster
size and tolerates very high-latency communication. Furthermore, it is easy to
deploy and use with no parameter tuning, and it is compatible with existing
Caffe models. We quantify the dependence of the speedup obtained by SparkNet on
the number of machines, the communication frequency, and the cluster's
communication overhead, and we benchmark our system's performance on the
ImageNet dataset.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 03:29:56 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Nov 2015 10:35:40 GMT"
},
{
"version": "v3",
"created": "Wed, 6 Jan 2016 07:48:06 GMT"
},
{
"version": "v4",
"created": "Sun, 28 Feb 2016 23:43:36 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Moritz",
"Philipp",
""
],
[
"Nishihara",
"Robert",
""
],
[
"Stoica",
"Ion",
""
],
[
"Jordan",
"Michael I.",
""
]
] | TITLE: SparkNet: Training Deep Networks in Spark
ABSTRACT: Training deep networks is a time-consuming process, with networks for object
recognition often requiring multiple days to train. For this reason, leveraging
the resources of a cluster to speed up training is an important area of work.
However, widely-popular batch-processing computational frameworks like
MapReduce and Spark were not designed to support the asynchronous and
communication-intensive workloads of existing distributed deep learning
systems. We introduce SparkNet, a framework for training deep networks in
Spark. Our implementation includes a convenient interface for reading data from
Spark RDDs, a Scala interface to the Caffe deep learning framework, and a
lightweight multi-dimensional tensor library. Using a simple parallelization
scheme for stochastic gradient descent, SparkNet scales well with the cluster
size and tolerates very high-latency communication. Furthermore, it is easy to
deploy and use with no parameter tuning, and it is compatible with existing
Caffe models. We quantify the dependence of the speedup obtained by SparkNet on
the number of machines, the communication frequency, and the cluster's
communication overhead, and we benchmark our system's performance on the
ImageNet dataset.
| no_new_dataset | 0.937669 |
1602.06245 | Paul Bendich | Paul Bendich, Ellen Gasparovic, Christopher J. Tralie, John Harer | Scaffoldings and Spines: Organizing High-Dimensional Data Using Cover
Trees, Local Principal Component Analysis, and Persistent Homology | 14 pages | null | null | null | cs.CG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a flexible and multi-scale method for organizing, visualizing, and
understanding datasets sampled from or near stratified spaces. The first part
of the algorithm produces a cover tree using adaptive thresholds based on a
combination of multi-scale local principal component analysis and topological
data analysis. The resulting cover tree nodes consist of points within or near
the same stratum of the stratified space. They are then connected to form a
\emph{scaffolding} graph, which is then simplified and collapsed down into a
\emph{spine} graph. From this latter graph the stratified structure becomes
apparent. We demonstrate our technique on several synthetic point cloud
examples and we use it to understand song structure in musical audio data.
| [
{
"version": "v1",
"created": "Fri, 19 Feb 2016 18:19:05 GMT"
},
{
"version": "v2",
"created": "Sat, 27 Feb 2016 05:21:51 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Bendich",
"Paul",
""
],
[
"Gasparovic",
"Ellen",
""
],
[
"Tralie",
"Christopher J.",
""
],
[
"Harer",
"John",
""
]
] | TITLE: Scaffoldings and Spines: Organizing High-Dimensional Data Using Cover
Trees, Local Principal Component Analysis, and Persistent Homology
ABSTRACT: We propose a flexible and multi-scale method for organizing, visualizing, and
understanding datasets sampled from or near stratified spaces. The first part
of the algorithm produces a cover tree using adaptive thresholds based on a
combination of multi-scale local principal component analysis and topological
data analysis. The resulting cover tree nodes consist of points within or near
the same stratum of the stratified space. They are then connected to form a
\emph{scaffolding} graph, which is then simplified and collapsed down into a
\emph{spine} graph. From this latter graph the stratified structure becomes
apparent. We demonstrate our technique on several synthetic point cloud
examples and we use it to understand song structure in musical audio data.
| no_new_dataset | 0.956675 |
1602.06550 | Igor Melnyk | Igor Melnyk, Arindam Banerjee, Bryan Matthews, and Nikunj Oza | Semi-Markov Switching Vector Autoregressive Model-based Anomaly
Detection in Aviation Systems | null | null | null | null | cs.LG stat.AP stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work we consider the problem of anomaly detection in heterogeneous,
multivariate, variable-length time series datasets. Our focus is on the
aviation safety domain, where data objects are flights and time series are
sensor readings and pilot switches. In this context the goal is to detect
anomalous flight segments, due to mechanical, environmental, or human factors
in order to identifying operationally significant events and provide insights
into the flight operations and highlight otherwise unavailable potential safety
risks and precursors to accidents. For this purpose, we propose a framework
which represents each flight using a semi-Markov switching vector
autoregressive (SMS-VAR) model. Detection of anomalies is then based on
measuring dissimilarities between the model's prediction and data observation.
The framework is scalable, due to the inherent parallel nature of most
computations, and can be used to perform online anomaly detection. Extensive
experimental results on simulated and real datasets illustrate that the
framework can detect various types of anomalies along with the key parameters
involved.
| [
{
"version": "v1",
"created": "Sun, 21 Feb 2016 16:55:36 GMT"
},
{
"version": "v2",
"created": "Sun, 28 Feb 2016 23:12:31 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Melnyk",
"Igor",
""
],
[
"Banerjee",
"Arindam",
""
],
[
"Matthews",
"Bryan",
""
],
[
"Oza",
"Nikunj",
""
]
] | TITLE: Semi-Markov Switching Vector Autoregressive Model-based Anomaly
Detection in Aviation Systems
ABSTRACT: In this work we consider the problem of anomaly detection in heterogeneous,
multivariate, variable-length time series datasets. Our focus is on the
aviation safety domain, where data objects are flights and time series are
sensor readings and pilot switches. In this context the goal is to detect
anomalous flight segments, due to mechanical, environmental, or human factors
in order to identifying operationally significant events and provide insights
into the flight operations and highlight otherwise unavailable potential safety
risks and precursors to accidents. For this purpose, we propose a framework
which represents each flight using a semi-Markov switching vector
autoregressive (SMS-VAR) model. Detection of anomalies is then based on
measuring dissimilarities between the model's prediction and data observation.
The framework is scalable, due to the inherent parallel nature of most
computations, and can be used to perform online anomaly detection. Extensive
experimental results on simulated and real datasets illustrate that the
framework can detect various types of anomalies along with the key parameters
involved.
| no_new_dataset | 0.943764 |
1602.08721 | Oren Kalinsky | Oren Kalinsky, Yoav Etsion, Benny Kimelfeld | Flexible Caching in Trie Joins | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional algorithms for multiway join computation are based on rewriting
the order of joins and combining results of intermediate subqueries. Recently,
several approaches have been proposed for algorithms that are "worst-case
optimal" wherein all relations are scanned simultaneously. An example is
Veldhuizen's Leapfrog Trie Join (LFTJ). An important advantage of LFTJ is its
small memory footprint, due to the fact that intermediate results are full
tuples that can be dumped immediately. However, since the algorithm does not
store intermediate results, recurring joins must be reconstructed from the
source relations, resulting in excessive memory traffic. In this paper, we
address this problem by incorporating caches into LFTJ. We do so by adopting
recent developments on join optimization, tying variable ordering to tree
decomposition. While the traditional usage of tree decomposition computes the
result for each bag in advance, our proposed approach incorporates caching
directly into LFTJ and can dynamically adjust the size of the cache.
Consequently, our solution balances memory usage and repeated computation, as
confirmed by our experiments over SNAP datasets.
| [
{
"version": "v1",
"created": "Sun, 28 Feb 2016 14:26:08 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Kalinsky",
"Oren",
""
],
[
"Etsion",
"Yoav",
""
],
[
"Kimelfeld",
"Benny",
""
]
] | TITLE: Flexible Caching in Trie Joins
ABSTRACT: Traditional algorithms for multiway join computation are based on rewriting
the order of joins and combining results of intermediate subqueries. Recently,
several approaches have been proposed for algorithms that are "worst-case
optimal" wherein all relations are scanned simultaneously. An example is
Veldhuizen's Leapfrog Trie Join (LFTJ). An important advantage of LFTJ is its
small memory footprint, due to the fact that intermediate results are full
tuples that can be dumped immediately. However, since the algorithm does not
store intermediate results, recurring joins must be reconstructed from the
source relations, resulting in excessive memory traffic. In this paper, we
address this problem by incorporating caches into LFTJ. We do so by adopting
recent developments on join optimization, tying variable ordering to tree
decomposition. While the traditional usage of tree decomposition computes the
result for each bag in advance, our proposed approach incorporates caching
directly into LFTJ and can dynamically adjust the size of the cache.
Consequently, our solution balances memory usage and repeated computation, as
confirmed by our experiments over SNAP datasets.
| no_new_dataset | 0.940298 |
1602.08791 | Vijay Gadepally | Vijay Gadepally, Jennie Duggan, Aaron Elmore, Jeremy Kepner, Samuel
Madden, Tim Mattson, Michael Stonebraker | The BigDAWG Architecture | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | BigDAWG is a polystore system designed to work on complex problems that
naturally span across different processing or storage engines. BigDAWG provides
an architecture that supports diverse database systems working with different
data models, support for the competing notions of location transparency and
semantic completeness via islands of information and a middleware that provides
a uniform multi-island interface. In this article, we describe the current
architecture of BigDAWG, its application on the MIMIC II medical dataset, and
our plans for the mechanics of cross-system queries. During the presentation,
we will also deliver a brief demonstration of the current version of BigDAWG.
| [
{
"version": "v1",
"created": "Mon, 29 Feb 2016 00:49:11 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Gadepally",
"Vijay",
""
],
[
"Duggan",
"Jennie",
""
],
[
"Elmore",
"Aaron",
""
],
[
"Kepner",
"Jeremy",
""
],
[
"Madden",
"Samuel",
""
],
[
"Mattson",
"Tim",
""
],
[
"Stonebraker",
"Michael",
""
]
] | TITLE: The BigDAWG Architecture
ABSTRACT: BigDAWG is a polystore system designed to work on complex problems that
naturally span across different processing or storage engines. BigDAWG provides
an architecture that supports diverse database systems working with different
data models, support for the competing notions of location transparency and
semantic completeness via islands of information and a middleware that provides
a uniform multi-island interface. In this article, we describe the current
architecture of BigDAWG, its application on the MIMIC II medical dataset, and
our plans for the mechanics of cross-system queries. During the presentation,
we will also deliver a brief demonstration of the current version of BigDAWG.
| no_new_dataset | 0.950088 |
1602.08855 | Corneliu Florea | Corneliu Florea, Razvan Condorovici, Constantin Vertan, Raluca Boia,
Laura Florea, Ruxandra Vranceanu | Pandora: Description of a Painting Database for Art Movement Recognition
with Baselines and Perspectives | 11 pages, 1 figure, 6 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To facilitate computer analysis of visual art, in the form of paintings, we
introduce Pandora (Paintings Dataset for Recognizing the Art movement)
database, a collection of digitized paintings labelled with respect to the
artistic movement. Noting that the set of databases available as benchmarks for
evaluation is highly reduced and most existing ones are limited in variability
and number of images, we propose a novel large scale dataset of digital
paintings. The database consists of more than 7700 images from 12 art
movements. Each genre is illustrated by a number of images varying from 250 to
nearly 1000. We investigate how local and global features and classification
systems are able to recognize the art movement. Our experimental results
suggest that accurate recognition is achievable by a combination of various
categories.To facilitate computer analysis of visual art, in the form of
paintings, we introduce Pandora (Paintings Dataset for Recognizing the Art
movement) database, a collection of digitized paintings labelled with respect
to the artistic movement. Noting that the set of databases available as
benchmarks for evaluation is highly reduced and most existing ones are limited
in variability and number of images, we propose a novel large scale dataset of
digital paintings. The database consists of more than 7700 images from 12 art
movements. Each genre is illustrated by a number of images varying from 250 to
nearly 1000. We investigate how local and global features and classification
systems are able to recognize the art movement. Our experimental results
suggest that accurate recognition is achievable by a combination of various
categories.
| [
{
"version": "v1",
"created": "Mon, 29 Feb 2016 08:24:01 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Florea",
"Corneliu",
""
],
[
"Condorovici",
"Razvan",
""
],
[
"Vertan",
"Constantin",
""
],
[
"Boia",
"Raluca",
""
],
[
"Florea",
"Laura",
""
],
[
"Vranceanu",
"Ruxandra",
""
]
] | TITLE: Pandora: Description of a Painting Database for Art Movement Recognition
with Baselines and Perspectives
ABSTRACT: To facilitate computer analysis of visual art, in the form of paintings, we
introduce Pandora (Paintings Dataset for Recognizing the Art movement)
database, a collection of digitized paintings labelled with respect to the
artistic movement. Noting that the set of databases available as benchmarks for
evaluation is highly reduced and most existing ones are limited in variability
and number of images, we propose a novel large scale dataset of digital
paintings. The database consists of more than 7700 images from 12 art
movements. Each genre is illustrated by a number of images varying from 250 to
nearly 1000. We investigate how local and global features and classification
systems are able to recognize the art movement. Our experimental results
suggest that accurate recognition is achievable by a combination of various
categories.To facilitate computer analysis of visual art, in the form of
paintings, we introduce Pandora (Paintings Dataset for Recognizing the Art
movement) database, a collection of digitized paintings labelled with respect
to the artistic movement. Noting that the set of databases available as
benchmarks for evaluation is highly reduced and most existing ones are limited
in variability and number of images, we propose a novel large scale dataset of
digital paintings. The database consists of more than 7700 images from 12 art
movements. Each genre is illustrated by a number of images varying from 250 to
nearly 1000. We investigate how local and global features and classification
systems are able to recognize the art movement. Our experimental results
suggest that accurate recognition is achievable by a combination of various
categories.
| new_dataset | 0.964589 |
1602.08978 | Yoshiharu Maeno | Yoshiharu Maeno | Epidemiological geographic profiling for a meta-population network | null | null | null | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Epidemiological geographic profiling is a statistical method for making
inferences about likely areas of a source from the geographical distribution of
patients. Epidemiological geographic profiling algorithms are developed to
locate a source from the dataset on the number of new cases for a
meta-population network model. It is found from the WHO dataset on the SARS
outbreak that Hong Kong remains the most likely source throughout the period of
observation. This reasoning is pertinent under the restricted circumstance that
the number of reported probable cases in China was missing, unreliable, and
incomprehensive. It may also imply that globally connected Hong Kong was more
influential as a spreader than China. Singapore, Taiwan, Canada, and the United
States follow Hong Kong in the likeliness ranking list.
| [
{
"version": "v1",
"created": "Tue, 17 Nov 2015 07:29:56 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Maeno",
"Yoshiharu",
""
]
] | TITLE: Epidemiological geographic profiling for a meta-population network
ABSTRACT: Epidemiological geographic profiling is a statistical method for making
inferences about likely areas of a source from the geographical distribution of
patients. Epidemiological geographic profiling algorithms are developed to
locate a source from the dataset on the number of new cases for a
meta-population network model. It is found from the WHO dataset on the SARS
outbreak that Hong Kong remains the most likely source throughout the period of
observation. This reasoning is pertinent under the restricted circumstance that
the number of reported probable cases in China was missing, unreliable, and
incomprehensive. It may also imply that globally connected Hong Kong was more
influential as a spreader than China. Singapore, Taiwan, Canada, and the United
States follow Hong Kong in the likeliness ranking list.
| no_new_dataset | 0.95096 |
1602.08986 | G\'eraud Le Falher | G\'eraud Le Falher and Fabio Vitale | Even Trolls Are Useful: Efficient Link Classification in Signed Networks | 17 pages, 3 figures | null | null | null | cs.LG cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We address the problem of classifying the links of signed social networks
given their full structural topology. Motivated by a binary user behaviour
assumption, which is supported by decades of research in psychology, we develop
an efficient and surprisingly simple approach to solve this classification
problem. Our methods operate both within the active and batch settings. We
demonstrate that the algorithms we developed are extremely fast in both
theoretical and practical terms. Within the active setting, we provide a new
complexity measure and a rigorous analysis of our methods that hold for
arbitrary signed networks. We validate our theoretical claims carrying out a
set of experiments on three well known real-world datasets, showing that our
methods outperform the competitors while being much faster.
| [
{
"version": "v1",
"created": "Mon, 29 Feb 2016 14:45:18 GMT"
}
] | 2016-03-01T00:00:00 | [
[
"Falher",
"Géraud Le",
""
],
[
"Vitale",
"Fabio",
""
]
] | TITLE: Even Trolls Are Useful: Efficient Link Classification in Signed Networks
ABSTRACT: We address the problem of classifying the links of signed social networks
given their full structural topology. Motivated by a binary user behaviour
assumption, which is supported by decades of research in psychology, we develop
an efficient and surprisingly simple approach to solve this classification
problem. Our methods operate both within the active and batch settings. We
demonstrate that the algorithms we developed are extremely fast in both
theoretical and practical terms. Within the active setting, we provide a new
complexity measure and a rigorous analysis of our methods that hold for
arbitrary signed networks. We validate our theoretical claims carrying out a
set of experiments on three well known real-world datasets, showing that our
methods outperform the competitors while being much faster.
| no_new_dataset | 0.944485 |
1510.03009 | Zhouhan Lin | Zhouhan Lin, Matthieu Courbariaux, Roland Memisevic, Yoshua Bengio | Neural Networks with Few Multiplications | Published as a conference paper at ICLR 2016. 9 pages, 3 figures | null | null | null | cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For most deep learning algorithms training is notoriously time consuming.
Since most of the computation in training neural networks is typically spent on
floating point multiplications, we investigate an approach to training that
eliminates the need for most of these. Our method consists of two parts: First
we stochastically binarize weights to convert multiplications involved in
computing hidden states to sign changes. Second, while back-propagating error
derivatives, in addition to binarizing the weights, we quantize the
representations at each layer to convert the remaining multiplications into
binary shifts. Experimental results across 3 popular datasets (MNIST, CIFAR10,
SVHN) show that this approach not only does not hurt classification performance
but can result in even better performance than standard stochastic gradient
descent training, paving the way to fast, hardware-friendly training of neural
networks.
| [
{
"version": "v1",
"created": "Sun, 11 Oct 2015 04:32:39 GMT"
},
{
"version": "v2",
"created": "Mon, 9 Nov 2015 20:16:10 GMT"
},
{
"version": "v3",
"created": "Fri, 26 Feb 2016 05:24:30 GMT"
}
] | 2016-02-29T00:00:00 | [
[
"Lin",
"Zhouhan",
""
],
[
"Courbariaux",
"Matthieu",
""
],
[
"Memisevic",
"Roland",
""
],
[
"Bengio",
"Yoshua",
""
]
] | TITLE: Neural Networks with Few Multiplications
ABSTRACT: For most deep learning algorithms training is notoriously time consuming.
Since most of the computation in training neural networks is typically spent on
floating point multiplications, we investigate an approach to training that
eliminates the need for most of these. Our method consists of two parts: First
we stochastically binarize weights to convert multiplications involved in
computing hidden states to sign changes. Second, while back-propagating error
derivatives, in addition to binarizing the weights, we quantize the
representations at each layer to convert the remaining multiplications into
binary shifts. Experimental results across 3 popular datasets (MNIST, CIFAR10,
SVHN) show that this approach not only does not hurt classification performance
but can result in even better performance than standard stochastic gradient
descent training, paving the way to fast, hardware-friendly training of neural
networks.
| no_new_dataset | 0.949153 |
1511.06426 | Moontae Lee | Moontae Lee, Xiaodong He, Wen-tau Yih, Jianfeng Gao, Li Deng, Paul
Smolensky | Reasoning in Vector Space: An Exploratory Study of Question Answering | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Question answering tasks have shown remarkable progress with distributed
vector representation. In this paper, we investigate the recently proposed
Facebook bAbI tasks which consist of twenty different categories of questions
that require complex reasoning. Because the previous work on bAbI are all
end-to-end models, errors could come from either an imperfect understanding of
semantics or in certain steps of the reasoning. For clearer analysis, we
propose two vector space models inspired by Tensor Product Representation (TPR)
to perform knowledge encoding and logical reasoning based on common-sense
inference. They together achieve near-perfect accuracy on all categories
including positional reasoning and path finding that have proved difficult for
most of the previous approaches. We hypothesize that the difficulties in these
categories are due to the multi-relations in contrast to uni-relational
characteristic of other categories. Our exploration sheds light on designing
more sophisticated dataset and moving one step toward integrating transparent
and interpretable formalism of TPR into existing learning paradigms.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 22:30:10 GMT"
},
{
"version": "v2",
"created": "Thu, 7 Jan 2016 22:30:01 GMT"
},
{
"version": "v3",
"created": "Tue, 19 Jan 2016 11:16:46 GMT"
},
{
"version": "v4",
"created": "Fri, 26 Feb 2016 18:49:34 GMT"
}
] | 2016-02-29T00:00:00 | [
[
"Lee",
"Moontae",
""
],
[
"He",
"Xiaodong",
""
],
[
"Yih",
"Wen-tau",
""
],
[
"Gao",
"Jianfeng",
""
],
[
"Deng",
"Li",
""
],
[
"Smolensky",
"Paul",
""
]
] | TITLE: Reasoning in Vector Space: An Exploratory Study of Question Answering
ABSTRACT: Question answering tasks have shown remarkable progress with distributed
vector representation. In this paper, we investigate the recently proposed
Facebook bAbI tasks which consist of twenty different categories of questions
that require complex reasoning. Because the previous work on bAbI are all
end-to-end models, errors could come from either an imperfect understanding of
semantics or in certain steps of the reasoning. For clearer analysis, we
propose two vector space models inspired by Tensor Product Representation (TPR)
to perform knowledge encoding and logical reasoning based on common-sense
inference. They together achieve near-perfect accuracy on all categories
including positional reasoning and path finding that have proved difficult for
most of the previous approaches. We hypothesize that the difficulties in these
categories are due to the multi-relations in contrast to uni-relational
characteristic of other categories. Our exploration sheds light on designing
more sophisticated dataset and moving one step toward integrating transparent
and interpretable formalism of TPR into existing learning paradigms.
| new_dataset | 0.968768 |
1602.08141 | Thomas Castelli | Thomas Castelli, Aidean Sharghi, Don Harper, Alain Tremeau and Mubarak
Shah | Autonomous navigation for low-altitude UAVs in urban areas | null | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, consumer Unmanned Aerial Vehicles have become very popular,
everyone can buy and fly a drone without previous experience, which raises
concern in regards to regulations and public safety. In this paper, we present
a novel approach towards enabling safe operation of such vehicles in urban
areas. Our method uses geodetically accurate dataset images with Geographical
Information System (GIS) data of road networks and buildings provided by Google
Maps, to compute a weighted A* shortest path from start to end locations of a
mission. Weights represent the potential risk of injuries for individuals in
all categories of land-use, i.e. flying over buildings is considered safer than
above roads. We enable safe UAV operation in regards to 1- land-use by
computing a static global path dependent on environmental structures, and 2-
avoiding flying over moving objects such as cars and pedestrians by dynamically
optimizing the path locally during the flight. As all input sources are first
geo-registered, pixels and GPS coordinates are equivalent, it therefore allows
us to generate an automated and user-friendly mission with GPS waypoints
readable by consumer drones' autopilots. We simulated 54 missions and show
significant improvement in maximizing UAV's standoff distance to moving objects
with a quantified safety parameter over 40 times better than the naive straight
line navigation.
| [
{
"version": "v1",
"created": "Thu, 25 Feb 2016 22:43:14 GMT"
}
] | 2016-02-29T00:00:00 | [
[
"Castelli",
"Thomas",
""
],
[
"Sharghi",
"Aidean",
""
],
[
"Harper",
"Don",
""
],
[
"Tremeau",
"Alain",
""
],
[
"Shah",
"Mubarak",
""
]
] | TITLE: Autonomous navigation for low-altitude UAVs in urban areas
ABSTRACT: In recent years, consumer Unmanned Aerial Vehicles have become very popular,
everyone can buy and fly a drone without previous experience, which raises
concern in regards to regulations and public safety. In this paper, we present
a novel approach towards enabling safe operation of such vehicles in urban
areas. Our method uses geodetically accurate dataset images with Geographical
Information System (GIS) data of road networks and buildings provided by Google
Maps, to compute a weighted A* shortest path from start to end locations of a
mission. Weights represent the potential risk of injuries for individuals in
all categories of land-use, i.e. flying over buildings is considered safer than
above roads. We enable safe UAV operation in regards to 1- land-use by
computing a static global path dependent on environmental structures, and 2-
avoiding flying over moving objects such as cars and pedestrians by dynamically
optimizing the path locally during the flight. As all input sources are first
geo-registered, pixels and GPS coordinates are equivalent, it therefore allows
us to generate an automated and user-friendly mission with GPS waypoints
readable by consumer drones' autopilots. We simulated 54 missions and show
significant improvement in maximizing UAV's standoff distance to moving objects
with a quantified safety parameter over 40 times better than the naive straight
line navigation.
| no_new_dataset | 0.947284 |
1602.08225 | Wei Liu | Wei Liu, Wei-Long Zheng, Bao-Liang Lu | Multimodal Emotion Recognition Using Multimodal Deep Learning | null | null | null | null | cs.HC cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To enhance the performance of affective models and reduce the cost of
acquiring physiological signals for real-world applications, we adopt
multimodal deep learning approach to construct affective models from multiple
physiological signals. For unimodal enhancement task, we indicate that the best
recognition accuracy of 82.11% on SEED dataset is achieved with shared
representations generated by Deep AutoEncoder (DAE) model. For multimodal
facilitation tasks, we demonstrate that the Bimodal Deep AutoEncoder (BDAE)
achieves the mean accuracies of 91.01% and 83.25% on SEED and DEAP datasets,
respectively, which are much superior to the state-of-the-art approaches. For
cross-modal learning task, our experimental results demonstrate that the mean
accuracy of 66.34% is achieved on SEED dataset through shared representations
generated by EEG-based DAE as training samples and shared representations
generated by eye-based DAE as testing sample, and vice versa.
| [
{
"version": "v1",
"created": "Fri, 26 Feb 2016 07:43:14 GMT"
}
] | 2016-02-29T00:00:00 | [
[
"Liu",
"Wei",
""
],
[
"Zheng",
"Wei-Long",
""
],
[
"Lu",
"Bao-Liang",
""
]
] | TITLE: Multimodal Emotion Recognition Using Multimodal Deep Learning
ABSTRACT: To enhance the performance of affective models and reduce the cost of
acquiring physiological signals for real-world applications, we adopt
multimodal deep learning approach to construct affective models from multiple
physiological signals. For unimodal enhancement task, we indicate that the best
recognition accuracy of 82.11% on SEED dataset is achieved with shared
representations generated by Deep AutoEncoder (DAE) model. For multimodal
facilitation tasks, we demonstrate that the Bimodal Deep AutoEncoder (BDAE)
achieves the mean accuracies of 91.01% and 83.25% on SEED and DEAP datasets,
respectively, which are much superior to the state-of-the-art approaches. For
cross-modal learning task, our experimental results demonstrate that the mean
accuracy of 66.34% is achieved on SEED dataset through shared representations
generated by EEG-based DAE as training samples and shared representations
generated by eye-based DAE as testing sample, and vice versa.
| no_new_dataset | 0.951504 |
1602.08393 | Anshumali Shrivastava | Anshumali Shrivastava | Exact Weighted Minwise Hashing in Constant Time | null | null | null | null | cs.DS cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Weighted minwise hashing (WMH) is one of the fundamental subroutine, required
by many celebrated approximation algorithms, commonly adopted in industrial
practice for large scale-search and learning. The resource bottleneck of the
algorithms is the computation of multiple (typically a few hundreds to
thousands) independent hashes of the data. The fastest hashing algorithm is by
Ioffe \cite{Proc:Ioffe_ICDM10}, which requires one pass over the entire data
vector, $O(d)$ ($d$ is the number of non-zeros), for computing one hash.
However, the requirement of multiple hashes demands hundreds or thousands
passes over the data. This is very costly for modern massive dataset.
In this work, we break this expensive barrier and show an expected constant
amortized time algorithm which computes $k$ independent and unbiased WMH in
time $O(k)$ instead of $O(dk)$ required by Ioffe's method. Moreover, our
proposal only needs a few bits (5 - 9 bits) of storage per hash value compared
to around $64$ bits required by the state-of-art-methodologies. Experimental
evaluations, on real datasets, show that for computing 500 WMH, our proposal
can be 60000x faster than the Ioffe's method without losing any accuracy. Our
method is also around 100x faster than approximate heuristics capitalizing on
the efficient "densified" one permutation hashing schemes
\cite{Proc:OneHashLSH_ICML14}. Given the simplicity of our approach and its
significant advantages, we hope that it will replace existing implementations
in practice.
| [
{
"version": "v1",
"created": "Fri, 26 Feb 2016 16:55:08 GMT"
}
] | 2016-02-29T00:00:00 | [
[
"Shrivastava",
"Anshumali",
""
]
] | TITLE: Exact Weighted Minwise Hashing in Constant Time
ABSTRACT: Weighted minwise hashing (WMH) is one of the fundamental subroutine, required
by many celebrated approximation algorithms, commonly adopted in industrial
practice for large scale-search and learning. The resource bottleneck of the
algorithms is the computation of multiple (typically a few hundreds to
thousands) independent hashes of the data. The fastest hashing algorithm is by
Ioffe \cite{Proc:Ioffe_ICDM10}, which requires one pass over the entire data
vector, $O(d)$ ($d$ is the number of non-zeros), for computing one hash.
However, the requirement of multiple hashes demands hundreds or thousands
passes over the data. This is very costly for modern massive dataset.
In this work, we break this expensive barrier and show an expected constant
amortized time algorithm which computes $k$ independent and unbiased WMH in
time $O(k)$ instead of $O(dk)$ required by Ioffe's method. Moreover, our
proposal only needs a few bits (5 - 9 bits) of storage per hash value compared
to around $64$ bits required by the state-of-art-methodologies. Experimental
evaluations, on real datasets, show that for computing 500 WMH, our proposal
can be 60000x faster than the Ioffe's method without losing any accuracy. Our
method is also around 100x faster than approximate heuristics capitalizing on
the efficient "densified" one permutation hashing schemes
\cite{Proc:OneHashLSH_ICML14}. Given the simplicity of our approach and its
significant advantages, we hope that it will replace existing implementations
in practice.
| no_new_dataset | 0.944177 |
1602.08447 | Le Hoang Son | Mumtaz Ali, Nguyen Van Minh, Le Hoang Son | A Neutrosophic Recommender System for Medical Diagnosis Based on
Algebraic Neutrosophic Measures | Keywords: Medical diagnosis, neutrosophic set, neutrosophic
recommender system, non-linear regression model | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neutrosophic set has the ability to handle uncertain, incomplete,
inconsistent, indeterminate information in a more accurate way. In this paper,
we proposed a neutrosophic recommender system to predict the diseases based on
neutrosophic set which includes single-criterion neutrosophic recommender
system (SC-NRS) and multi-criterion neutrosophic recommender system (MC-NRS).
Further, we investigated some algebraic operations of neutrosophic recommender
system such as union, complement, intersection, probabilistic sum, bold sum,
bold intersection, bounded difference, symmetric difference, convex linear sum
of min and max operators, Cartesian product, associativity, commutativity and
distributive. Based on these operations, we studied the algebraic structures
such as lattices, Kleen algebra, de Morgan algebra, Brouwerian algebra, BCK
algebra, Stone algebra and MV algebra. In addition, we introduced several types
of similarity measures based on these algebraic operations and studied some of
their theoretic properties. Moreover, we accomplished a prediction formula
using the proposed algebraic similarity measure. We also proposed a new
algorithm for medical diagnosis based on neutrosophic recommender system.
Finally to check the validity of the proposed methodology, we made experiments
on the datasets Heart, RHC, Breast cancer, Diabetes and DMD. At the end, we
presented the MSE and computational time by comparing the proposed algorithm
with the relevant ones such as ICSM, DSM, CARE, CFMD, as well as other variants
namely Variant 67, Variant 69, and Varian 71 both in tabular and graphical form
to analyze the efficiency and accuracy. Finally we analyzed the strength of all
8 algorithms by ANOVA statistical tool.
| [
{
"version": "v1",
"created": "Thu, 25 Feb 2016 03:20:00 GMT"
}
] | 2016-02-29T00:00:00 | [
[
"Ali",
"Mumtaz",
""
],
[
"Van Minh",
"Nguyen",
""
],
[
"Son",
"Le Hoang",
""
]
] | TITLE: A Neutrosophic Recommender System for Medical Diagnosis Based on
Algebraic Neutrosophic Measures
ABSTRACT: Neutrosophic set has the ability to handle uncertain, incomplete,
inconsistent, indeterminate information in a more accurate way. In this paper,
we proposed a neutrosophic recommender system to predict the diseases based on
neutrosophic set which includes single-criterion neutrosophic recommender
system (SC-NRS) and multi-criterion neutrosophic recommender system (MC-NRS).
Further, we investigated some algebraic operations of neutrosophic recommender
system such as union, complement, intersection, probabilistic sum, bold sum,
bold intersection, bounded difference, symmetric difference, convex linear sum
of min and max operators, Cartesian product, associativity, commutativity and
distributive. Based on these operations, we studied the algebraic structures
such as lattices, Kleen algebra, de Morgan algebra, Brouwerian algebra, BCK
algebra, Stone algebra and MV algebra. In addition, we introduced several types
of similarity measures based on these algebraic operations and studied some of
their theoretic properties. Moreover, we accomplished a prediction formula
using the proposed algebraic similarity measure. We also proposed a new
algorithm for medical diagnosis based on neutrosophic recommender system.
Finally to check the validity of the proposed methodology, we made experiments
on the datasets Heart, RHC, Breast cancer, Diabetes and DMD. At the end, we
presented the MSE and computational time by comparing the proposed algorithm
with the relevant ones such as ICSM, DSM, CARE, CFMD, as well as other variants
namely Variant 67, Variant 69, and Varian 71 both in tabular and graphical form
to analyze the efficiency and accuracy. Finally we analyzed the strength of all
8 algorithms by ANOVA statistical tool.
| no_new_dataset | 0.945551 |
1508.06163 | Guangliang Cheng | Guangliang Cheng, Feiyun Zhu, Shiming Xiang and Chunhong Pan | Accurate Urban Road Centerline Extraction from VHR Imagery via
Multiscale Segmentation and Tensor Voting | 25 pages, 11 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | It is very useful and increasingly popular to extract accurate road
centerlines from very-high-resolution (VHR) re- mote sensing imagery for
various applications, such as road map generation and updating etc. There are
three shortcomings of current methods: (a) Due to the noise and occlusions
(owing to vehicles and trees), most road extraction methods bring in
heterogeneous classification results; (b) Morphological thinning algorithm is
widely used to extract road centerlines, while it pro- duces small spurs around
the centerlines; (c) Many methods are ineffective to extract centerlines around
the road intersections. To address the above three issues, we propose a novel
method to ex- tract smooth and complete road centerlines via three techniques:
the multiscale joint collaborative representation (MJCR) & graph cuts (GC),
tensor voting (TV) & non-maximum suppression (NMS) and fitting based connection
algorithm. Specifically, a MJCR-GC based road area segmentation method is
proposed by incorporating mutiscale features and spatial information. In this
way, a homogenous road segmentation result is achieved. Then, to obtain a
smooth and correct road centerline network, a TV-NMS based centerline
extraction method is introduced. This method not only extracts smooth road
centerlines, but also connects the discontinuous road centerlines. Finally, to
overcome the ineffectiveness of current methods in the road intersection, a
fitting based road centerline connection algorithm is proposed. As a result, we
can get a complete road centerline network. Extensive experiments on two
datasets demonstrate that our method achieves higher quantitative results, as
well as more satisfactory visual performances by comparing with state-of-the-
art methods.
| [
{
"version": "v1",
"created": "Tue, 25 Aug 2015 14:18:34 GMT"
},
{
"version": "v2",
"created": "Thu, 25 Feb 2016 15:29:25 GMT"
}
] | 2016-02-26T00:00:00 | [
[
"Cheng",
"Guangliang",
""
],
[
"Zhu",
"Feiyun",
""
],
[
"Xiang",
"Shiming",
""
],
[
"Pan",
"Chunhong",
""
]
] | TITLE: Accurate Urban Road Centerline Extraction from VHR Imagery via
Multiscale Segmentation and Tensor Voting
ABSTRACT: It is very useful and increasingly popular to extract accurate road
centerlines from very-high-resolution (VHR) re- mote sensing imagery for
various applications, such as road map generation and updating etc. There are
three shortcomings of current methods: (a) Due to the noise and occlusions
(owing to vehicles and trees), most road extraction methods bring in
heterogeneous classification results; (b) Morphological thinning algorithm is
widely used to extract road centerlines, while it pro- duces small spurs around
the centerlines; (c) Many methods are ineffective to extract centerlines around
the road intersections. To address the above three issues, we propose a novel
method to ex- tract smooth and complete road centerlines via three techniques:
the multiscale joint collaborative representation (MJCR) & graph cuts (GC),
tensor voting (TV) & non-maximum suppression (NMS) and fitting based connection
algorithm. Specifically, a MJCR-GC based road area segmentation method is
proposed by incorporating mutiscale features and spatial information. In this
way, a homogenous road segmentation result is achieved. Then, to obtain a
smooth and correct road centerline network, a TV-NMS based centerline
extraction method is introduced. This method not only extracts smooth road
centerlines, but also connects the discontinuous road centerlines. Finally, to
overcome the ineffectiveness of current methods in the road intersection, a
fitting based road centerline connection algorithm is proposed. As a result, we
can get a complete road centerline network. Extensive experiments on two
datasets demonstrate that our method achieves higher quantitative results, as
well as more satisfactory visual performances by comparing with state-of-the-
art methods.
| no_new_dataset | 0.954774 |
1602.06844 | Hao Wu | Hao Wu, Yue Ning, Prithwish Chakraborty, Jilles Vreeken, Nikolaj Tatti
and Naren Ramakrishnan | Generating Realistic Synthetic Population Datasets | The conference version of the paper is submitted for publication | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern studies of societal phenomena rely on the availability of large
datasets capturing attributes and activities of synthetic, city-level,
populations. For instance, in epidemiology, synthetic population datasets are
necessary to study disease propagation and intervention measures before
implementation. In social science, synthetic population datasets are needed to
understand how policy decisions might affect preferences and behaviors of
individuals. In public health, synthetic population datasets are necessary to
capture diagnostic and procedural characteristics of patient records without
violating confidentialities of individuals. To generate such datasets over a
large set of categorical variables, we propose the use of the maximum entropy
principle to formalize a generative model such that in a statistically
well-founded way we can optimally utilize given prior information about the
data, and are unbiased otherwise. An efficient inference algorithm is designed
to estimate the maximum entropy model, and we demonstrate how our approach is
adept at estimating underlying data distributions. We evaluate this approach
against both simulated data and on US census datasets, and demonstrate its
feasibility using an epidemic simulation application.
| [
{
"version": "v1",
"created": "Mon, 22 Feb 2016 16:28:17 GMT"
},
{
"version": "v2",
"created": "Tue, 23 Feb 2016 16:50:26 GMT"
},
{
"version": "v3",
"created": "Thu, 25 Feb 2016 17:20:42 GMT"
}
] | 2016-02-26T00:00:00 | [
[
"Wu",
"Hao",
""
],
[
"Ning",
"Yue",
""
],
[
"Chakraborty",
"Prithwish",
""
],
[
"Vreeken",
"Jilles",
""
],
[
"Tatti",
"Nikolaj",
""
],
[
"Ramakrishnan",
"Naren",
""
]
] | TITLE: Generating Realistic Synthetic Population Datasets
ABSTRACT: Modern studies of societal phenomena rely on the availability of large
datasets capturing attributes and activities of synthetic, city-level,
populations. For instance, in epidemiology, synthetic population datasets are
necessary to study disease propagation and intervention measures before
implementation. In social science, synthetic population datasets are needed to
understand how policy decisions might affect preferences and behaviors of
individuals. In public health, synthetic population datasets are necessary to
capture diagnostic and procedural characteristics of patient records without
violating confidentialities of individuals. To generate such datasets over a
large set of categorical variables, we propose the use of the maximum entropy
principle to formalize a generative model such that in a statistically
well-founded way we can optimally utilize given prior information about the
data, and are unbiased otherwise. An efficient inference algorithm is designed
to estimate the maximum entropy model, and we demonstrate how our approach is
adept at estimating underlying data distributions. We evaluate this approach
against both simulated data and on US census datasets, and demonstrate its
feasibility using an epidemic simulation application.
| no_new_dataset | 0.868994 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.