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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1606.02492 | Shreyas Saxena | Shreyas Saxena and Jakob Verbeek | Convolutional Neural Fabrics | Corrected typos (In proceedings of NIPS16 ) | null | null | null | cs.CV cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the success of CNNs, selecting the optimal architecture for a given
task remains an open problem. Instead of aiming to select a single optimal
architecture, we propose a "fabric" that embeds an exponentially large number
of architectures. The fabric consists of a 3D trellis that connects response
maps at different layers, scales, and channels with a sparse homogeneous local
connectivity pattern. The only hyper-parameters of a fabric are the number of
channels and layers. While individual architectures can be recovered as paths,
the fabric can in addition ensemble all embedded architectures together,
sharing their weights where their paths overlap. Parameters can be learned
using standard methods based on back-propagation, at a cost that scales
linearly in the fabric size. We present benchmark results competitive with the
state of the art for image classification on MNIST and CIFAR10, and for
semantic segmentation on the Part Labels dataset.
| [
{
"version": "v1",
"created": "Wed, 8 Jun 2016 10:17:51 GMT"
},
{
"version": "v2",
"created": "Thu, 9 Jun 2016 16:21:57 GMT"
},
{
"version": "v3",
"created": "Fri, 28 Oct 2016 13:10:05 GMT"
},
{
"version": "v4",
"created": "Mon, 30 Jan 2017 12:28:29 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Saxena",
"Shreyas",
""
],
[
"Verbeek",
"Jakob",
""
]
] | TITLE: Convolutional Neural Fabrics
ABSTRACT: Despite the success of CNNs, selecting the optimal architecture for a given
task remains an open problem. Instead of aiming to select a single optimal
architecture, we propose a "fabric" that embeds an exponentially large number
of architectures. The fabric consists of a 3D trellis that connects response
maps at different layers, scales, and channels with a sparse homogeneous local
connectivity pattern. The only hyper-parameters of a fabric are the number of
channels and layers. While individual architectures can be recovered as paths,
the fabric can in addition ensemble all embedded architectures together,
sharing their weights where their paths overlap. Parameters can be learned
using standard methods based on back-propagation, at a cost that scales
linearly in the fabric size. We present benchmark results competitive with the
state of the art for image classification on MNIST and CIFAR10, and for
semantic segmentation on the Part Labels dataset.
| no_new_dataset | 0.946349 |
1607.04673 | Abhineet Singh | Abhineet Singh, Mennatullah Siam and Martin Jagersand | Unifying Registration based Tracking: A Case Study with Structural
Similarity | Accepted at WACV 2017. Supplementary available at:
http://webdocs.cs.ualberta.ca/~vis/mtf/ssim_supplementary.pdf arXiv admin
note: text overlap with arXiv:1603.01292 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper adapts a popular image quality measure called structural
similarity for high precision registration based tracking while also
introducing a simpler and faster variant of the same. Further, these are
evaluated comprehensively against existing measures using a unified approach to
study registration based trackers that decomposes them into three constituent
sub modules - appearance model, state space model and search method. Several
popular trackers in literature are broken down using this method so that their
contributions - as of this paper - are shown to be limited to only one or two
of these submodules. An open source tracking framework is made available that
follows this decomposition closely through extensive use of generic
programming. It is used to perform all experiments on four publicly available
datasets so the results are easily reproducible. This framework provides a
convenient interface to plug in a new method for any sub module and combine it
with existing methods for the other two. It can also serve as a fast and
flexible solution for practical tracking needs due to its highly efficient
implementation.
| [
{
"version": "v1",
"created": "Fri, 15 Jul 2016 22:25:46 GMT"
},
{
"version": "v2",
"created": "Fri, 7 Oct 2016 08:19:18 GMT"
},
{
"version": "v3",
"created": "Mon, 17 Oct 2016 04:52:14 GMT"
},
{
"version": "v4",
"created": "Mon, 30 Jan 2017 14:50:49 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Singh",
"Abhineet",
""
],
[
"Siam",
"Mennatullah",
""
],
[
"Jagersand",
"Martin",
""
]
] | TITLE: Unifying Registration based Tracking: A Case Study with Structural
Similarity
ABSTRACT: This paper adapts a popular image quality measure called structural
similarity for high precision registration based tracking while also
introducing a simpler and faster variant of the same. Further, these are
evaluated comprehensively against existing measures using a unified approach to
study registration based trackers that decomposes them into three constituent
sub modules - appearance model, state space model and search method. Several
popular trackers in literature are broken down using this method so that their
contributions - as of this paper - are shown to be limited to only one or two
of these submodules. An open source tracking framework is made available that
follows this decomposition closely through extensive use of generic
programming. It is used to perform all experiments on four publicly available
datasets so the results are easily reproducible. This framework provides a
convenient interface to plug in a new method for any sub module and combine it
with existing methods for the other two. It can also serve as a fast and
flexible solution for practical tracking needs due to its highly efficient
implementation.
| no_new_dataset | 0.939692 |
1610.05861 | Samarth Manoj Brahmbhatt | Samarth Brahmbhatt, Henrik I. Christensen and James Hays | StuffNet: Using 'Stuff' to Improve Object Detection | Camera-ready version for IEEE WACV 2017 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a Convolutional Neural Network (CNN) based algorithm - StuffNet -
for object detection. In addition to the standard convolutional features
trained for region proposal and object detection [31], StuffNet uses
convolutional features trained for segmentation of objects and 'stuff'
(amorphous categories such as ground and water). Through experiments on Pascal
VOC 2010, we show the importance of features learnt from stuff segmentation for
improving object detection performance. StuffNet improves performance from
18.8% mAP to 23.9% mAP for small objects. We also devise a method to train
StuffNet on datasets that do not have stuff segmentation labels. Through
experiments on Pascal VOC 2007 and 2012, we demonstrate the effectiveness of
this method and show that StuffNet also significantly improves object detection
performance on such datasets.
| [
{
"version": "v1",
"created": "Wed, 19 Oct 2016 04:44:51 GMT"
},
{
"version": "v2",
"created": "Mon, 30 Jan 2017 03:10:20 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Brahmbhatt",
"Samarth",
""
],
[
"Christensen",
"Henrik I.",
""
],
[
"Hays",
"James",
""
]
] | TITLE: StuffNet: Using 'Stuff' to Improve Object Detection
ABSTRACT: We propose a Convolutional Neural Network (CNN) based algorithm - StuffNet -
for object detection. In addition to the standard convolutional features
trained for region proposal and object detection [31], StuffNet uses
convolutional features trained for segmentation of objects and 'stuff'
(amorphous categories such as ground and water). Through experiments on Pascal
VOC 2010, we show the importance of features learnt from stuff segmentation for
improving object detection performance. StuffNet improves performance from
18.8% mAP to 23.9% mAP for small objects. We also devise a method to train
StuffNet on datasets that do not have stuff segmentation labels. Through
experiments on Pascal VOC 2007 and 2012, we demonstrate the effectiveness of
this method and show that StuffNet also significantly improves object detection
performance on such datasets.
| no_new_dataset | 0.955361 |
1612.05079 | Ankur Handa | John McCormac, Ankur Handa, Stefan Leutenegger, Andrew J. Davison | SceneNet RGB-D: 5M Photorealistic Images of Synthetic Indoor
Trajectories with Ground Truth | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce SceneNet RGB-D, expanding the previous work of SceneNet to
enable large scale photorealistic rendering of indoor scene trajectories. It
provides pixel-perfect ground truth for scene understanding problems such as
semantic segmentation, instance segmentation, and object detection, and also
for geometric computer vision problems such as optical flow, depth estimation,
camera pose estimation, and 3D reconstruction. Random sampling permits
virtually unlimited scene configurations, and here we provide a set of 5M
rendered RGB-D images from over 15K trajectories in synthetic layouts with
random but physically simulated object poses. Each layout also has random
lighting, camera trajectories, and textures. The scale of this dataset is well
suited for pre-training data-driven computer vision techniques from scratch
with RGB-D inputs, which previously has been limited by relatively small
labelled datasets in NYUv2 and SUN RGB-D. It also provides a basis for
investigating 3D scene labelling tasks by providing perfect camera poses and
depth data as proxy for a SLAM system. We host the dataset at
http://robotvault.bitbucket.io/scenenet-rgbd.html
| [
{
"version": "v1",
"created": "Thu, 15 Dec 2016 14:22:38 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Dec 2016 01:37:54 GMT"
},
{
"version": "v3",
"created": "Mon, 30 Jan 2017 11:06:14 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"McCormac",
"John",
""
],
[
"Handa",
"Ankur",
""
],
[
"Leutenegger",
"Stefan",
""
],
[
"Davison",
"Andrew J.",
""
]
] | TITLE: SceneNet RGB-D: 5M Photorealistic Images of Synthetic Indoor
Trajectories with Ground Truth
ABSTRACT: We introduce SceneNet RGB-D, expanding the previous work of SceneNet to
enable large scale photorealistic rendering of indoor scene trajectories. It
provides pixel-perfect ground truth for scene understanding problems such as
semantic segmentation, instance segmentation, and object detection, and also
for geometric computer vision problems such as optical flow, depth estimation,
camera pose estimation, and 3D reconstruction. Random sampling permits
virtually unlimited scene configurations, and here we provide a set of 5M
rendered RGB-D images from over 15K trajectories in synthetic layouts with
random but physically simulated object poses. Each layout also has random
lighting, camera trajectories, and textures. The scale of this dataset is well
suited for pre-training data-driven computer vision techniques from scratch
with RGB-D inputs, which previously has been limited by relatively small
labelled datasets in NYUv2 and SUN RGB-D. It also provides a basis for
investigating 3D scene labelling tasks by providing perfect camera poses and
depth data as proxy for a SLAM system. We host the dataset at
http://robotvault.bitbucket.io/scenenet-rgbd.html
| new_dataset | 0.955068 |
1701.07368 | Zhenzhong Lan | Zhenzhong Lan, Yi Zhu, Alexander G. Hauptmann | Deep Local Video Feature for Action Recognition | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the problem of representing an entire video using CNN features
for human action recognition. Currently, limited by GPU memory, we have not
been able to feed a whole video into CNN/RNNs for end-to-end learning. A common
practice is to use sampled frames as inputs and video labels as supervision.
One major problem of this popular approach is that the local samples may not
contain the information indicated by global labels. To deal with this problem,
we propose to treat the deep networks trained on local inputs as local feature
extractors. After extracting local features, we aggregate them into global
features and train another mapping function on the same training data to map
the global features into global labels. We study a set of problems regarding
this new type of local features such as how to aggregate them into global
features. Experimental results on HMDB51 and UCF101 datasets show that, for
these new local features, a simple maximum pooling on the sparsely sampled
features lead to significant performance improvement.
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 16:23:17 GMT"
},
{
"version": "v2",
"created": "Sat, 28 Jan 2017 13:50:09 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Lan",
"Zhenzhong",
""
],
[
"Zhu",
"Yi",
""
],
[
"Hauptmann",
"Alexander G.",
""
]
] | TITLE: Deep Local Video Feature for Action Recognition
ABSTRACT: We investigate the problem of representing an entire video using CNN features
for human action recognition. Currently, limited by GPU memory, we have not
been able to feed a whole video into CNN/RNNs for end-to-end learning. A common
practice is to use sampled frames as inputs and video labels as supervision.
One major problem of this popular approach is that the local samples may not
contain the information indicated by global labels. To deal with this problem,
we propose to treat the deep networks trained on local inputs as local feature
extractors. After extracting local features, we aggregate them into global
features and train another mapping function on the same training data to map
the global features into global labels. We study a set of problems regarding
this new type of local features such as how to aggregate them into global
features. Experimental results on HMDB51 and UCF101 datasets show that, for
these new local features, a simple maximum pooling on the sparsely sampled
features lead to significant performance improvement.
| no_new_dataset | 0.944177 |
1701.08241 | Yansong Gao | Yansong Gao, Hua Ma, Geifei Li, Shaza Zeitouni, Said F. Al-Sarawi,
Derek Abbott, Ahmad-Reza Sadeghi, Damith C. Ranasinghe | Exploiting PUF Models for Error Free Response Generation | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Physical unclonable functions (PUF) extract secrets from randomness inherent
in manufacturing processes. PUFs are utilized for basic cryptographic tasks
such as authentication and key generation, and more recently, to realize key
exchange and bit commitment requiring a large number of error free responses
from a strong PUF. We propose an approach to eliminate the need to implement
expensive on-chip error correction logic implementation and the associated
helper data storage to reconcile naturally noisy PUF responses. In particular,
we exploit a statistical model of an Arbiter PUF (APUF) constructed under the
nominal operating condition during the challenge response enrollment phase by a
trusted party to judiciously select challenges that yield error-free responses
even across a wide operating conditions, specifically, a $ \pm 20\% $ supply
voltage variation and a $ 40^{\crc} $ temperature variation. We validate our
approach using measurements from two APUF datasets. Experimental results
indicate that large number of error-free responses can be generated on demand
under worst-case when PUF response error rate is up to 16.68\%.
| [
{
"version": "v1",
"created": "Sat, 28 Jan 2017 03:06:33 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Gao",
"Yansong",
""
],
[
"Ma",
"Hua",
""
],
[
"Li",
"Geifei",
""
],
[
"Zeitouni",
"Shaza",
""
],
[
"Al-Sarawi",
"Said F.",
""
],
[
"Abbott",
"Derek",
""
],
[
"Sadeghi",
"Ahmad-Reza",
""
],
[
"Ranasinghe",
"Damith C.",
""
]
] | TITLE: Exploiting PUF Models for Error Free Response Generation
ABSTRACT: Physical unclonable functions (PUF) extract secrets from randomness inherent
in manufacturing processes. PUFs are utilized for basic cryptographic tasks
such as authentication and key generation, and more recently, to realize key
exchange and bit commitment requiring a large number of error free responses
from a strong PUF. We propose an approach to eliminate the need to implement
expensive on-chip error correction logic implementation and the associated
helper data storage to reconcile naturally noisy PUF responses. In particular,
we exploit a statistical model of an Arbiter PUF (APUF) constructed under the
nominal operating condition during the challenge response enrollment phase by a
trusted party to judiciously select challenges that yield error-free responses
even across a wide operating conditions, specifically, a $ \pm 20\% $ supply
voltage variation and a $ 40^{\crc} $ temperature variation. We validate our
approach using measurements from two APUF datasets. Experimental results
indicate that large number of error-free responses can be generated on demand
under worst-case when PUF response error rate is up to 16.68\%.
| no_new_dataset | 0.947039 |
1701.08291 | Ilke Cugu | \.Ilke \c{C}u\u{g}u, Eren \c{S}ener, \c{C}a\u{g}r{\i} Erciyes, Burak
Balc{\i}, Emre Ak{\i}n, It{\i}r \"Onal, Ahmet O\u{g}uz Aky\"uz | Treelogy: A Novel Tree Classifier Utilizing Deep and Hand-crafted
Representations | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We propose a novel tree classification system called Treelogy, that fuses
deep representations with hand-crafted features obtained from leaf images to
perform leaf-based plant classification. Key to this system are segmentation of
the leaf from an untextured background, using convolutional neural networks
(CNNs) for learning deep representations, extracting hand-crafted features with
a number of image processing techniques, training a linear SVM with feature
vectors, merging SVM and CNN results, and identifying the species from a
dataset of 57 trees. Our classification results show that fusion of deep
representations with hand-crafted features leads to the highest accuracy. The
proposed algorithm is embedded in a smart-phone application, which is publicly
available. Furthermore, our novel dataset comprised of 5408 leaf images is also
made public for use of other researchers.
| [
{
"version": "v1",
"created": "Sat, 28 Jan 2017 13:41:49 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Çuğu",
"İlke",
""
],
[
"Şener",
"Eren",
""
],
[
"Erciyes",
"Çağrı",
""
],
[
"Balcı",
"Burak",
""
],
[
"Akın",
"Emre",
""
],
[
"Önal",
"Itır",
""
],
[
"Akyüz",
"Ahmet Oğuz",
""
]
] | TITLE: Treelogy: A Novel Tree Classifier Utilizing Deep and Hand-crafted
Representations
ABSTRACT: We propose a novel tree classification system called Treelogy, that fuses
deep representations with hand-crafted features obtained from leaf images to
perform leaf-based plant classification. Key to this system are segmentation of
the leaf from an untextured background, using convolutional neural networks
(CNNs) for learning deep representations, extracting hand-crafted features with
a number of image processing techniques, training a linear SVM with feature
vectors, merging SVM and CNN results, and identifying the species from a
dataset of 57 trees. Our classification results show that fusion of deep
representations with hand-crafted features leads to the highest accuracy. The
proposed algorithm is embedded in a smart-phone application, which is publicly
available. Furthermore, our novel dataset comprised of 5408 leaf images is also
made public for use of other researchers.
| new_dataset | 0.95877 |
1701.08318 | Xueliang (Leon) Liu | Xueliang Liu | Deep Recurrent Neural Network for Protein Function Prediction from
Sequence | null | null | null | null | q-bio.QM cs.LG q-bio.BM stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As high-throughput biological sequencing becomes faster and cheaper, the need
to extract useful information from sequencing becomes ever more paramount,
often limited by low-throughput experimental characterizations. For proteins,
accurate prediction of their functions directly from their primary amino-acid
sequences has been a long standing challenge. Here, machine learning using
artificial recurrent neural networks (RNN) was applied towards classification
of protein function directly from primary sequence without sequence alignment,
heuristic scoring or feature engineering. The RNN models containing
long-short-term-memory (LSTM) units trained on public, annotated datasets from
UniProt achieved high performance for in-class prediction of four important
protein functions tested, particularly compared to other machine learning
algorithms using sequence-derived protein features. RNN models were used also
for out-of-class predictions of phylogenetically distinct protein families with
similar functions, including proteins of the CRISPR-associated nuclease,
ferritin-like iron storage and cytochrome P450 families. Applying the trained
RNN models on the partially unannotated UniRef100 database predicted not only
candidates validated by existing annotations but also currently unannotated
sequences. Some RNN predictions for the ferritin-like iron sequestering
function were experimentally validated, even though their sequences differ
significantly from known, characterized proteins and from each other and cannot
be easily predicted using popular bioinformatics methods. As sequencing and
experimental characterization data increases rapidly, the machine-learning
approach based on RNN could be useful for discovery and prediction of
homologues for a wide range of protein functions.
| [
{
"version": "v1",
"created": "Sat, 28 Jan 2017 19:33:59 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Liu",
"Xueliang",
""
]
] | TITLE: Deep Recurrent Neural Network for Protein Function Prediction from
Sequence
ABSTRACT: As high-throughput biological sequencing becomes faster and cheaper, the need
to extract useful information from sequencing becomes ever more paramount,
often limited by low-throughput experimental characterizations. For proteins,
accurate prediction of their functions directly from their primary amino-acid
sequences has been a long standing challenge. Here, machine learning using
artificial recurrent neural networks (RNN) was applied towards classification
of protein function directly from primary sequence without sequence alignment,
heuristic scoring or feature engineering. The RNN models containing
long-short-term-memory (LSTM) units trained on public, annotated datasets from
UniProt achieved high performance for in-class prediction of four important
protein functions tested, particularly compared to other machine learning
algorithms using sequence-derived protein features. RNN models were used also
for out-of-class predictions of phylogenetically distinct protein families with
similar functions, including proteins of the CRISPR-associated nuclease,
ferritin-like iron storage and cytochrome P450 families. Applying the trained
RNN models on the partially unannotated UniRef100 database predicted not only
candidates validated by existing annotations but also currently unannotated
sequences. Some RNN predictions for the ferritin-like iron sequestering
function were experimentally validated, even though their sequences differ
significantly from known, characterized proteins and from each other and cannot
be easily predicted using popular bioinformatics methods. As sequencing and
experimental characterization data increases rapidly, the machine-learning
approach based on RNN could be useful for discovery and prediction of
homologues for a wide range of protein functions.
| no_new_dataset | 0.951142 |
1701.08380 | Martin Thoma | Martin Thoma | The HASYv2 dataset | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper describes the HASYv2 dataset. HASY is a publicly available, free
of charge dataset of single symbols similar to MNIST. It contains 168233
instances of 369 classes. HASY contains two challenges: A classification
challenge with 10 pre-defined folds for 10-fold cross-validation and a
verification challenge.
| [
{
"version": "v1",
"created": "Sun, 29 Jan 2017 13:42:14 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Thoma",
"Martin",
""
]
] | TITLE: The HASYv2 dataset
ABSTRACT: This paper describes the HASYv2 dataset. HASY is a publicly available, free
of charge dataset of single symbols similar to MNIST. It contains 168233
instances of 369 classes. HASY contains two challenges: A classification
challenge with 10 pre-defined folds for 10-fold cross-validation and a
verification challenge.
| new_dataset | 0.959875 |
1701.08694 | Saiful Islam Md | Md. Saiful Islam, Fazla Elahi Md Jubayer and Syed Ikhtiar Ahmed | A Comparative Study on Different Types of Approaches to Bengali document
Categorization | 6 pages | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Document categorization is a technique where the category of a document is
determined. In this paper three well-known supervised learning techniques which
are Support Vector Machine(SVM), Na\"ive Bayes(NB) and Stochastic Gradient
Descent(SGD) compared for Bengali document categorization. Besides classifier,
classification also depends on how feature is selected from dataset. For
analyzing those classifier performances on predicting a document against twelve
categories several feature selection techniques are also applied in this
article namely Chi square distribution, normalized TFIDF (term
frequency-inverse document frequency) with word analyzer. So, we attempt to
explore the efficiency of those three-classification algorithms by using two
different feature selection techniques in this article.
| [
{
"version": "v1",
"created": "Fri, 27 Jan 2017 13:08:08 GMT"
}
] | 2017-01-31T00:00:00 | [
[
"Islam",
"Md. Saiful",
""
],
[
"Jubayer",
"Fazla Elahi Md",
""
],
[
"Ahmed",
"Syed Ikhtiar",
""
]
] | TITLE: A Comparative Study on Different Types of Approaches to Bengali document
Categorization
ABSTRACT: Document categorization is a technique where the category of a document is
determined. In this paper three well-known supervised learning techniques which
are Support Vector Machine(SVM), Na\"ive Bayes(NB) and Stochastic Gradient
Descent(SGD) compared for Bengali document categorization. Besides classifier,
classification also depends on how feature is selected from dataset. For
analyzing those classifier performances on predicting a document against twelve
categories several feature selection techniques are also applied in this
article namely Chi square distribution, normalized TFIDF (term
frequency-inverse document frequency) with word analyzer. So, we attempt to
explore the efficiency of those three-classification algorithms by using two
different feature selection techniques in this article.
| no_new_dataset | 0.953362 |
1604.03489 | Xavier Gir\'o-i-Nieto | Victor Campos, Brendan Jou and Xavier Giro-i-Nieto | From Pixels to Sentiment: Fine-tuning CNNs for Visual Sentiment
Prediction | Accepted for publication in Image and Vision Computing. Models and
source code available at https://github.com/imatge-upc/sentiment-2016 | null | null | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visual multimedia have become an inseparable part of our digital social
lives, and they often capture moments tied with deep affections. Automated
visual sentiment analysis tools can provide a means of extracting the rich
feelings and latent dispositions embedded in these media. In this work, we
explore how Convolutional Neural Networks (CNNs), a now de facto computational
machine learning tool particularly in the area of Computer Vision, can be
specifically applied to the task of visual sentiment prediction. We accomplish
this through fine-tuning experiments using a state-of-the-art CNN and via
rigorous architecture analysis, we present several modifications that lead to
accuracy improvements over prior art on a dataset of images from a popular
social media platform. We additionally present visualizations of local patterns
that the network learned to associate with image sentiment for insight into how
visual positivity (or negativity) is perceived by the model.
| [
{
"version": "v1",
"created": "Tue, 12 Apr 2016 17:24:39 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Jan 2017 18:02:16 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Campos",
"Victor",
""
],
[
"Jou",
"Brendan",
""
],
[
"Giro-i-Nieto",
"Xavier",
""
]
] | TITLE: From Pixels to Sentiment: Fine-tuning CNNs for Visual Sentiment
Prediction
ABSTRACT: Visual multimedia have become an inseparable part of our digital social
lives, and they often capture moments tied with deep affections. Automated
visual sentiment analysis tools can provide a means of extracting the rich
feelings and latent dispositions embedded in these media. In this work, we
explore how Convolutional Neural Networks (CNNs), a now de facto computational
machine learning tool particularly in the area of Computer Vision, can be
specifically applied to the task of visual sentiment prediction. We accomplish
this through fine-tuning experiments using a state-of-the-art CNN and via
rigorous architecture analysis, we present several modifications that lead to
accuracy improvements over prior art on a dataset of images from a popular
social media platform. We additionally present visualizations of local patterns
that the network learned to associate with image sentiment for insight into how
visual positivity (or negativity) is perceived by the model.
| no_new_dataset | 0.946547 |
1608.07102 | Qiang Liu | Qiang Liu, Shu Wu, Liang Wang | Multi-behavioral Sequential Prediction with Recurrent Log-bilinear Model | IEEE Transactions on Knowledge and Data Engineering (TKDE), to appear | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid growth of Internet applications, sequential prediction in
collaborative filtering has become an emerging and crucial task. Given the
behavioral history of a specific user, predicting his or her next choice plays
a key role in improving various online services. Meanwhile, there are more and
more scenarios with multiple types of behaviors, while existing works mainly
study sequences with a single type of behavior. As a widely used approach,
Markov chain based models are based on a strong independence assumption. As two
classical neural network methods for modeling sequences, recurrent neural
networks cannot well model short-term contexts, and the log-bilinear model is
not suitable for long-term contexts. In this paper, we propose a Recurrent
Log-BiLinear (RLBL) model. It can model multiple types of behaviors in
historical sequences with behavior-specific transition matrices. RLBL applies a
recurrent structure for modeling long-term contexts. It models several items in
each hidden layer and employs position-specific transition matrices for
modeling short-term contexts. Moreover, considering continuous time difference
in behavioral history is a key factor for dynamic prediction, we further extend
RLBL and replace position-specific transition matrices with time-specific
transition matrices, and accordingly propose a Time-Aware Recurrent
Log-BiLinear (TA-RLBL) model. Experimental results show that the proposed RLBL
model and TA-RLBL model yield significant improvements over the competitive
compared methods on three datasets, i.e., Movielens-1M dataset, Global
Terrorism Database and Tmall dataset with different numbers of behavior types.
| [
{
"version": "v1",
"created": "Thu, 25 Aug 2016 12:01:18 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Oct 2016 09:08:40 GMT"
},
{
"version": "v3",
"created": "Thu, 8 Dec 2016 06:58:18 GMT"
},
{
"version": "v4",
"created": "Fri, 27 Jan 2017 09:53:14 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Liu",
"Qiang",
""
],
[
"Wu",
"Shu",
""
],
[
"Wang",
"Liang",
""
]
] | TITLE: Multi-behavioral Sequential Prediction with Recurrent Log-bilinear Model
ABSTRACT: With the rapid growth of Internet applications, sequential prediction in
collaborative filtering has become an emerging and crucial task. Given the
behavioral history of a specific user, predicting his or her next choice plays
a key role in improving various online services. Meanwhile, there are more and
more scenarios with multiple types of behaviors, while existing works mainly
study sequences with a single type of behavior. As a widely used approach,
Markov chain based models are based on a strong independence assumption. As two
classical neural network methods for modeling sequences, recurrent neural
networks cannot well model short-term contexts, and the log-bilinear model is
not suitable for long-term contexts. In this paper, we propose a Recurrent
Log-BiLinear (RLBL) model. It can model multiple types of behaviors in
historical sequences with behavior-specific transition matrices. RLBL applies a
recurrent structure for modeling long-term contexts. It models several items in
each hidden layer and employs position-specific transition matrices for
modeling short-term contexts. Moreover, considering continuous time difference
in behavioral history is a key factor for dynamic prediction, we further extend
RLBL and replace position-specific transition matrices with time-specific
transition matrices, and accordingly propose a Time-Aware Recurrent
Log-BiLinear (TA-RLBL) model. Experimental results show that the proposed RLBL
model and TA-RLBL model yield significant improvements over the competitive
compared methods on three datasets, i.e., Movielens-1M dataset, Global
Terrorism Database and Tmall dataset with different numbers of behavior types.
| no_new_dataset | 0.953579 |
1611.04741 | K. M. Annervaz | Biswajit Paria, K. M. Annervaz, Ambedkar Dukkipati, Ankush Chatterjee,
Sanjay Podder | A Neural Architecture Mimicking Humans End-to-End for Natural Language
Inference | 8 pages, 2 figures | null | null | null | cs.CL | http://creativecommons.org/publicdomain/zero/1.0/ | In this work we use the recent advances in representation learning to propose
a neural architecture for the problem of natural language inference. Our
approach is aligned to mimic how a human does the natural language inference
process given two statements. The model uses variants of Long Short Term Memory
(LSTM), attention mechanism and composable neural networks, to carry out the
task. Each part of our model can be mapped to a clear functionality humans do
for carrying out the overall task of natural language inference. The model is
end-to-end differentiable enabling training by stochastic gradient descent. On
Stanford Natural Language Inference(SNLI) dataset, the proposed model achieves
better accuracy numbers than all published models in literature.
| [
{
"version": "v1",
"created": "Tue, 15 Nov 2016 08:48:22 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Jan 2017 05:36:05 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Paria",
"Biswajit",
""
],
[
"Annervaz",
"K. M.",
""
],
[
"Dukkipati",
"Ambedkar",
""
],
[
"Chatterjee",
"Ankush",
""
],
[
"Podder",
"Sanjay",
""
]
] | TITLE: A Neural Architecture Mimicking Humans End-to-End for Natural Language
Inference
ABSTRACT: In this work we use the recent advances in representation learning to propose
a neural architecture for the problem of natural language inference. Our
approach is aligned to mimic how a human does the natural language inference
process given two statements. The model uses variants of Long Short Term Memory
(LSTM), attention mechanism and composable neural networks, to carry out the
task. Each part of our model can be mapped to a clear functionality humans do
for carrying out the overall task of natural language inference. The model is
end-to-end differentiable enabling training by stochastic gradient descent. On
Stanford Natural Language Inference(SNLI) dataset, the proposed model achieves
better accuracy numbers than all published models in literature.
| no_new_dataset | 0.951188 |
1611.09630 | Jakub Tomczak Ph.D. | Jakub M. Tomczak and Max Welling | Improving Variational Auto-Encoders using Householder Flow | A corrected version of the paper submitted to Bayesian Deep Learning
Workshop (NIPS 2016) | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Variational auto-encoders (VAE) are scalable and powerful generative models.
However, the choice of the variational posterior determines tractability and
flexibility of the VAE. Commonly, latent variables are modeled using the normal
distribution with a diagonal covariance matrix. This results in computational
efficiency but typically it is not flexible enough to match the true posterior
distribution. One fashion of enriching the variational posterior distribution
is application of normalizing flows, i.e., a series of invertible
transformations to latent variables with a simple posterior. In this paper, we
follow this line of thinking and propose a volume-preserving flow that uses a
series of Householder transformations. We show empirically on MNIST dataset and
histopathology data that the proposed flow allows to obtain more flexible
variational posterior and competitive results comparing to other normalizing
flows.
| [
{
"version": "v1",
"created": "Tue, 29 Nov 2016 13:49:31 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Dec 2016 12:04:28 GMT"
},
{
"version": "v3",
"created": "Sun, 22 Jan 2017 18:49:14 GMT"
},
{
"version": "v4",
"created": "Fri, 27 Jan 2017 00:36:51 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Tomczak",
"Jakub M.",
""
],
[
"Welling",
"Max",
""
]
] | TITLE: Improving Variational Auto-Encoders using Householder Flow
ABSTRACT: Variational auto-encoders (VAE) are scalable and powerful generative models.
However, the choice of the variational posterior determines tractability and
flexibility of the VAE. Commonly, latent variables are modeled using the normal
distribution with a diagonal covariance matrix. This results in computational
efficiency but typically it is not flexible enough to match the true posterior
distribution. One fashion of enriching the variational posterior distribution
is application of normalizing flows, i.e., a series of invertible
transformations to latent variables with a simple posterior. In this paper, we
follow this line of thinking and propose a volume-preserving flow that uses a
series of Householder transformations. We show empirically on MNIST dataset and
histopathology data that the proposed flow allows to obtain more flexible
variational posterior and competitive results comparing to other normalizing
flows.
| no_new_dataset | 0.950088 |
1701.07901 | Jingkuan Song Dr. | Jingkuan Song, Tao He, Lianli Gao, Xing Xu, Heng Tao Shen | Deep Region Hashing for Efficient Large-scale Instance Search from
Images | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Instance Search (INS) is a fundamental problem for many applications, while
it is more challenging comparing to traditional image search since the
relevancy is defined at the instance level.
Existing works have demonstrated the success of many complex ensemble systems
that are typically conducted by firstly generating object proposals, and then
extracting handcrafted and/or CNN features of each proposal for matching.
However, object bounding box proposals and feature extraction are often
conducted in two separated steps, thus the effectiveness of these methods
collapses. Also, due to the large amount of generated proposals, matching speed
becomes the bottleneck that limits its application to large-scale datasets. To
tackle these issues, in this paper we propose an effective and efficient Deep
Region Hashing (DRH) approach for large-scale INS using an image patch as the
query. Specifically, DRH is an end-to-end deep neural network which consists of
object proposal, feature extraction, and hash code generation. DRH shares
full-image convolutional feature map with the region proposal network, thus
enabling nearly cost-free region proposals. Also, each high-dimensional,
real-valued region features are mapped onto a low-dimensional, compact binary
codes for the efficient object region level matching on large-scale dataset.
Experimental results on four datasets show that our DRH can achieve even better
performance than the state-of-the-arts in terms of MAP, while the efficiency is
improved by nearly 100 times.
| [
{
"version": "v1",
"created": "Thu, 26 Jan 2017 23:18:58 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Song",
"Jingkuan",
""
],
[
"He",
"Tao",
""
],
[
"Gao",
"Lianli",
""
],
[
"Xu",
"Xing",
""
],
[
"Shen",
"Heng Tao",
""
]
] | TITLE: Deep Region Hashing for Efficient Large-scale Instance Search from
Images
ABSTRACT: Instance Search (INS) is a fundamental problem for many applications, while
it is more challenging comparing to traditional image search since the
relevancy is defined at the instance level.
Existing works have demonstrated the success of many complex ensemble systems
that are typically conducted by firstly generating object proposals, and then
extracting handcrafted and/or CNN features of each proposal for matching.
However, object bounding box proposals and feature extraction are often
conducted in two separated steps, thus the effectiveness of these methods
collapses. Also, due to the large amount of generated proposals, matching speed
becomes the bottleneck that limits its application to large-scale datasets. To
tackle these issues, in this paper we propose an effective and efficient Deep
Region Hashing (DRH) approach for large-scale INS using an image patch as the
query. Specifically, DRH is an end-to-end deep neural network which consists of
object proposal, feature extraction, and hash code generation. DRH shares
full-image convolutional feature map with the region proposal network, thus
enabling nearly cost-free region proposals. Also, each high-dimensional,
real-valued region features are mapped onto a low-dimensional, compact binary
codes for the efficient object region level matching on large-scale dataset.
Experimental results on four datasets show that our DRH can achieve even better
performance than the state-of-the-arts in terms of MAP, while the efficiency is
improved by nearly 100 times.
| no_new_dataset | 0.949059 |
1701.08022 | Sebastian Deorowicz | Marek Kokot and Maciej D{\l}ugosz and Sebastian Deorowicz | KMC 3: counting and manipulating k-mer statistics | null | null | null | null | q-bio.GN cs.DC cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Summary: Counting all k-mers in a given dataset is a standard procedure in
many bioinformatics applications. We introduce KMC3, a significant improvement
of the former KMC2 algorithm together with KMC tools for manipulating k-mer
databases. Usefulness of the tools is shown on a few real problems.
Availability: Program is freely available at
http://sun.aei.polsl.pl/REFRESH/kmc. Contact: [email protected]
| [
{
"version": "v1",
"created": "Fri, 27 Jan 2017 12:04:30 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Kokot",
"Marek",
""
],
[
"Długosz",
"Maciej",
""
],
[
"Deorowicz",
"Sebastian",
""
]
] | TITLE: KMC 3: counting and manipulating k-mer statistics
ABSTRACT: Summary: Counting all k-mers in a given dataset is a standard procedure in
many bioinformatics applications. We introduce KMC3, a significant improvement
of the former KMC2 algorithm together with KMC tools for manipulating k-mer
databases. Usefulness of the tools is shown on a few real problems.
Availability: Program is freely available at
http://sun.aei.polsl.pl/REFRESH/kmc. Contact: [email protected]
| no_new_dataset | 0.944791 |
1701.08128 | Gabriele Santi Mr | Gabriele Santi and Leonardo De Laurentiis | Evaluating a sublinear-time algorithm for the Minimum Spanning Tree
Weight problem | 23 pages, 13 figures, project developed during Master's Degree
studies | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present an implementation and an experimental evaluation of an algorithm
that, given a connected graph G (represented by adjacency lists), estimates in
sublinear time, with a relative error, the Minimum Spanning Tree Weight of G;
the original algorithm has been presented in "Approximating the minimum
spanning tree weight in sublinear time", by Bernard Chazelle, Ronitt Rubinfeld,
and Luca Trevisan (published with SIAM, DOI 10.1137/S0097539702403244). Since
the theoretical performances have already been shown and demonstrated in the
above-mentioned paper, our goal is, exclusively, to experimental evaluate the
algorithm and at last to present the results. Initially we discuss about some
theoretical aspects that arose while we were valuating the asymptotic
complexity of our specific implementation. Some technical insights are then
given on the implementation of the algorithm and on the dataset used in the
test phase, hence to show how the experiment has been carried out even for
reproducibility purposes; the results are then evaluated empirically and widely
discussed, comparing these with the performances of the Prim algorithm and the
Kruskal algorithm, launching several runs on a heterogeneous set of graphs and
different theoretical models for them.
| [
{
"version": "v1",
"created": "Fri, 27 Jan 2017 17:45:04 GMT"
}
] | 2017-01-30T00:00:00 | [
[
"Santi",
"Gabriele",
""
],
[
"De Laurentiis",
"Leonardo",
""
]
] | TITLE: Evaluating a sublinear-time algorithm for the Minimum Spanning Tree
Weight problem
ABSTRACT: We present an implementation and an experimental evaluation of an algorithm
that, given a connected graph G (represented by adjacency lists), estimates in
sublinear time, with a relative error, the Minimum Spanning Tree Weight of G;
the original algorithm has been presented in "Approximating the minimum
spanning tree weight in sublinear time", by Bernard Chazelle, Ronitt Rubinfeld,
and Luca Trevisan (published with SIAM, DOI 10.1137/S0097539702403244). Since
the theoretical performances have already been shown and demonstrated in the
above-mentioned paper, our goal is, exclusively, to experimental evaluate the
algorithm and at last to present the results. Initially we discuss about some
theoretical aspects that arose while we were valuating the asymptotic
complexity of our specific implementation. Some technical insights are then
given on the implementation of the algorithm and on the dataset used in the
test phase, hence to show how the experiment has been carried out even for
reproducibility purposes; the results are then evaluated empirically and widely
discussed, comparing these with the performances of the Prim algorithm and the
Kruskal algorithm, launching several runs on a heterogeneous set of graphs and
different theoretical models for them.
| no_new_dataset | 0.940463 |
1507.04921 | Chi Ho Yeung | Chi Ho Yeung | Do recommender systems benefit users? | 15 pages, 6 figures | J. Stat. Mech. 043401 (2016) | 10.1088/1742-5468/2016/04/043401 | null | cs.CY cs.IR cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recommender systems are present in many web applications to guide our
choices. They increase sales and benefit sellers, but whether they benefit
customers by providing relevant products is questionable. Here we introduce a
model to examine the benefit of recommender systems for users, and found that
recommendations from the system can be equivalent to random draws if one relies
too strongly on the system. Nevertheless, with sufficient information about
user preferences, recommendations become accurate and an abrupt transition to
this accurate regime is observed for some algorithms. On the other hand, we
found that a high accuracy evaluated by common accuracy metrics does not
necessarily correspond to a high real accuracy nor a benefit for users, which
serves as an alarm for operators and researchers of recommender systems. We
tested our model with a real dataset and observed similar behaviors. Finally, a
recommendation approach with improved accuracy is suggested. These results
imply that recommender systems can benefit users, but relying too strongly on
the system may render the system ineffective.
| [
{
"version": "v1",
"created": "Sun, 5 Jul 2015 14:56:11 GMT"
}
] | 2017-01-27T00:00:00 | [
[
"Yeung",
"Chi Ho",
""
]
] | TITLE: Do recommender systems benefit users?
ABSTRACT: Recommender systems are present in many web applications to guide our
choices. They increase sales and benefit sellers, but whether they benefit
customers by providing relevant products is questionable. Here we introduce a
model to examine the benefit of recommender systems for users, and found that
recommendations from the system can be equivalent to random draws if one relies
too strongly on the system. Nevertheless, with sufficient information about
user preferences, recommendations become accurate and an abrupt transition to
this accurate regime is observed for some algorithms. On the other hand, we
found that a high accuracy evaluated by common accuracy metrics does not
necessarily correspond to a high real accuracy nor a benefit for users, which
serves as an alarm for operators and researchers of recommender systems. We
tested our model with a real dataset and observed similar behaviors. Finally, a
recommendation approach with improved accuracy is suggested. These results
imply that recommender systems can benefit users, but relying too strongly on
the system may render the system ineffective.
| no_new_dataset | 0.949856 |
1701.07483 | Ashwin Venkataraman | Srikanth Jagabathula, Lakshminarayanan Subramanian, Ashwin
Venkataraman | A Model-based Projection Technique for Segmenting Customers | 51 pages, 3 figures, 4 tables | null | null | null | stat.ME cs.LG stat.AP stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of segmenting a large population of customers into
non-overlapping groups with similar preferences, using diverse preference
observations such as purchases, ratings, clicks, etc. over subsets of items. We
focus on the setting where the universe of items is large (ranging from
thousands to millions) and unstructured (lacking well-defined attributes) and
each customer provides observations for only a few items. These data
characteristics limit the applicability of existing techniques in marketing and
machine learning. To overcome these limitations, we propose a model-based
projection technique, which transforms the diverse set of observations into a
more comparable scale and deals with missing data by projecting the transformed
data onto a low-dimensional space. We then cluster the projected data to obtain
the customer segments. Theoretically, we derive precise necessary and
sufficient conditions that guarantee asymptotic recovery of the true customer
segments. Empirically, we demonstrate the speed and performance of our method
in two real-world case studies: (a) 84% improvement in the accuracy of new
movie recommendations on the MovieLens data set and (b) 6% improvement in the
performance of similar item recommendations algorithm on an offline dataset at
eBay. We show that our method outperforms standard latent-class and
demographic-based techniques.
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 20:47:40 GMT"
}
] | 2017-01-27T00:00:00 | [
[
"Jagabathula",
"Srikanth",
""
],
[
"Subramanian",
"Lakshminarayanan",
""
],
[
"Venkataraman",
"Ashwin",
""
]
] | TITLE: A Model-based Projection Technique for Segmenting Customers
ABSTRACT: We consider the problem of segmenting a large population of customers into
non-overlapping groups with similar preferences, using diverse preference
observations such as purchases, ratings, clicks, etc. over subsets of items. We
focus on the setting where the universe of items is large (ranging from
thousands to millions) and unstructured (lacking well-defined attributes) and
each customer provides observations for only a few items. These data
characteristics limit the applicability of existing techniques in marketing and
machine learning. To overcome these limitations, we propose a model-based
projection technique, which transforms the diverse set of observations into a
more comparable scale and deals with missing data by projecting the transformed
data onto a low-dimensional space. We then cluster the projected data to obtain
the customer segments. Theoretically, we derive precise necessary and
sufficient conditions that guarantee asymptotic recovery of the true customer
segments. Empirically, we demonstrate the speed and performance of our method
in two real-world case studies: (a) 84% improvement in the accuracy of new
movie recommendations on the MovieLens data set and (b) 6% improvement in the
performance of similar item recommendations algorithm on an offline dataset at
eBay. We show that our method outperforms standard latent-class and
demographic-based techniques.
| no_new_dataset | 0.948822 |
1701.07490 | RoopTeja Muppalla | Michele Miller, Dr. Tanvi Banerjee, RoopTeja Muppalla, Dr. William
Romine, Dr. Amit Sheth | What Are People Tweeting about Zika? An Exploratory Study Concerning
Symptoms, Treatment, Transmission, and Prevention | null | null | null | null | cs.SI q-bio.OT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The purpose of this study was to do a dataset distribution analysis, a
classification performance analysis, and a topical analysis concerning what
people are tweeting about four disease characteristics: symptoms, transmission,
prevention, and treatment. A combination of natural language processing and
machine learning techniques were used to determine what people are tweeting
about Zika. Specifically, a two-stage classifier system was built to find
relevant tweets on Zika, and then categorize these into the four disease
categories. Tweets in each disease category were then examined using latent
dirichlet allocation (LDA) to determine the five main tweet topics for each
disease characteristic. Results 1,234,605 tweets were collected. Tweets by
males and females were similar (28% and 23% respectively). The classifier
performed well on the training and test data for relevancy (F=0.87 and 0.99
respectively) and disease characteristics (F=0.79 and 0.90 respectively). Five
topics for each category were found and discussed with a focus on the symptoms
category. Through this process, we demonstrate how misinformation can be
discovered so that public health officials can respond to the tweets with
misinformation.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 18:52:22 GMT"
}
] | 2017-01-27T00:00:00 | [
[
"Miller",
"Michele",
""
],
[
"Banerjee",
"Dr. Tanvi",
""
],
[
"Muppalla",
"RoopTeja",
""
],
[
"Romine",
"Dr. William",
""
],
[
"Sheth",
"Dr. Amit",
""
]
] | TITLE: What Are People Tweeting about Zika? An Exploratory Study Concerning
Symptoms, Treatment, Transmission, and Prevention
ABSTRACT: The purpose of this study was to do a dataset distribution analysis, a
classification performance analysis, and a topical analysis concerning what
people are tweeting about four disease characteristics: symptoms, transmission,
prevention, and treatment. A combination of natural language processing and
machine learning techniques were used to determine what people are tweeting
about Zika. Specifically, a two-stage classifier system was built to find
relevant tweets on Zika, and then categorize these into the four disease
categories. Tweets in each disease category were then examined using latent
dirichlet allocation (LDA) to determine the five main tweet topics for each
disease characteristic. Results 1,234,605 tweets were collected. Tweets by
males and females were similar (28% and 23% respectively). The classifier
performed well on the training and test data for relevancy (F=0.87 and 0.99
respectively) and disease characteristics (F=0.79 and 0.90 respectively). Five
topics for each category were found and discussed with a focus on the symptoms
category. Through this process, we demonstrate how misinformation can be
discovered so that public health officials can respond to the tweets with
misinformation.
| no_new_dataset | 0.946646 |
1701.07732 | Liang Zheng | Liang Zheng, Yujia Huang, Huchuan Lu, Yi Yang | Pose Invariant Embedding for Deep Person Re-identification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pedestrian misalignment, which mainly arises from detector errors and pose
variations, is a critical problem for a robust person re-identification (re-ID)
system. With bad alignment, the background noise will significantly compromise
the feature learning and matching process. To address this problem, this paper
introduces the pose invariant embedding (PIE) as a pedestrian descriptor.
First, in order to align pedestrians to a standard pose, the PoseBox structure
is introduced, which is generated through pose estimation followed by affine
transformations. Second, to reduce the impact of pose estimation errors and
information loss during PoseBox construction, we design a PoseBox fusion (PBF)
CNN architecture that takes the original image, the PoseBox, and the pose
estimation confidence as input. The proposed PIE descriptor is thus defined as
the fully connected layer of the PBF network for the retrieval task.
Experiments are conducted on the Market-1501, CUHK03, and VIPeR datasets. We
show that PoseBox alone yields decent re-ID accuracy and that when integrated
in the PBF network, the learned PIE descriptor produces competitive performance
compared with the state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Thu, 26 Jan 2017 14:59:19 GMT"
}
] | 2017-01-27T00:00:00 | [
[
"Zheng",
"Liang",
""
],
[
"Huang",
"Yujia",
""
],
[
"Lu",
"Huchuan",
""
],
[
"Yang",
"Yi",
""
]
] | TITLE: Pose Invariant Embedding for Deep Person Re-identification
ABSTRACT: Pedestrian misalignment, which mainly arises from detector errors and pose
variations, is a critical problem for a robust person re-identification (re-ID)
system. With bad alignment, the background noise will significantly compromise
the feature learning and matching process. To address this problem, this paper
introduces the pose invariant embedding (PIE) as a pedestrian descriptor.
First, in order to align pedestrians to a standard pose, the PoseBox structure
is introduced, which is generated through pose estimation followed by affine
transformations. Second, to reduce the impact of pose estimation errors and
information loss during PoseBox construction, we design a PoseBox fusion (PBF)
CNN architecture that takes the original image, the PoseBox, and the pose
estimation confidence as input. The proposed PIE descriptor is thus defined as
the fully connected layer of the PBF network for the retrieval task.
Experiments are conducted on the Market-1501, CUHK03, and VIPeR datasets. We
show that PoseBox alone yields decent re-ID accuracy and that when integrated
in the PBF network, the learned PIE descriptor produces competitive performance
compared with the state-of-the-art approaches.
| no_new_dataset | 0.946892 |
1701.07773 | Bodhitha Jayatilaka | S. Amerio, S. Behari, J. Boyd, M. Brochmann, R. Culbertson, M.
Diesburg, J. Freeman, L. Garren, H. Greenlee, K. Herner, R. Illingworth, B.
Jayatilaka, A. Jonckheere, Q. Li, S. Naymola, G. Oleynik, W. Sakumotob, E.
Varnes, C. Vellidis, G. Watts, S. White | Data preservation at the Fermilab Tevatron | null | Nucl. Instrum. Methods Phys. Res. Sect. A, 851, 1 (2017) | 10.1016/j.nima.2017.01.043 | FERMILAB-PUB-16-552-CD | hep-ex physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Fermilab Tevatron collider's data-taking run ended in September 2011,
yielding a dataset with rich scientific potential. The CDF and D0 experiments
each have approximately 9 PB of collider and simulated data stored on tape. A
large computing infrastructure consisting of tape storage, disk cache, and
distributed grid computing for physics analysis with the Tevatron data is
present at Fermilab. The Fermilab Run II data preservation project intends to
keep this analysis capability sustained through the year 2020 and beyond. To
achieve this goal, we have implemented a system that utilizes virtualization,
automated validation, and migration to new standards in both software and data
storage technology and leverages resources available from currently-running
experiments at Fermilab. These efforts have also provided useful lessons in
ensuring long-term data access for numerous experiments, and enable
high-quality scientific output for years to come.
| [
{
"version": "v1",
"created": "Thu, 26 Jan 2017 16:54:34 GMT"
}
] | 2017-01-27T00:00:00 | [
[
"Amerio",
"S.",
""
],
[
"Behari",
"S.",
""
],
[
"Boyd",
"J.",
""
],
[
"Brochmann",
"M.",
""
],
[
"Culbertson",
"R.",
""
],
[
"Diesburg",
"M.",
""
],
[
"Freeman",
"J.",
""
],
[
"Garren",
"L.",
""
],
[
"Greenlee",
"H.",
""
],
[
"Herner",
"K.",
""
],
[
"Illingworth",
"R.",
""
],
[
"Jayatilaka",
"B.",
""
],
[
"Jonckheere",
"A.",
""
],
[
"Li",
"Q.",
""
],
[
"Naymola",
"S.",
""
],
[
"Oleynik",
"G.",
""
],
[
"Sakumotob",
"W.",
""
],
[
"Varnes",
"E.",
""
],
[
"Vellidis",
"C.",
""
],
[
"Watts",
"G.",
""
],
[
"White",
"S.",
""
]
] | TITLE: Data preservation at the Fermilab Tevatron
ABSTRACT: The Fermilab Tevatron collider's data-taking run ended in September 2011,
yielding a dataset with rich scientific potential. The CDF and D0 experiments
each have approximately 9 PB of collider and simulated data stored on tape. A
large computing infrastructure consisting of tape storage, disk cache, and
distributed grid computing for physics analysis with the Tevatron data is
present at Fermilab. The Fermilab Run II data preservation project intends to
keep this analysis capability sustained through the year 2020 and beyond. To
achieve this goal, we have implemented a system that utilizes virtualization,
automated validation, and migration to new standards in both software and data
storage technology and leverages resources available from currently-running
experiments at Fermilab. These efforts have also provided useful lessons in
ensuring long-term data access for numerous experiments, and enable
high-quality scientific output for years to come.
| no_new_dataset | 0.938294 |
1602.03468 | Abdolrahim Kadkhodamohammadi | Abdolrahim Kadkhodamohammadi, Afshin Gangi, Michel de Mathelin and
Nicolas Padoy | Articulated Clinician Detection Using 3D Pictorial Structures on RGB-D
Data | The supplementary video is available at https://youtu.be/iabbGSqRSgE | null | 10.1016/j.media.2016.07.001 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reliable human pose estimation (HPE) is essential to many clinical
applications, such as surgical workflow analysis, radiation safety monitoring
and human-robot cooperation. Proposed methods for the operating room (OR) rely
either on foreground estimation using a multi-camera system, which is a
challenge in real ORs due to color similarities and frequent illumination
changes, or on wearable sensors or markers, which are invasive and therefore
difficult to introduce in the room. Instead, we propose a novel approach based
on Pictorial Structures (PS) and on RGB-D data, which can be easily deployed in
real ORs. We extend the PS framework in two ways. First, we build robust and
discriminative part detectors using both color and depth images. We also
present a novel descriptor for depth images, called histogram of depth
differences (HDD). Second, we extend PS to 3D by proposing 3D pairwise
constraints and a new method that makes exact inference tractable. Our approach
is evaluated for pose estimation and clinician detection on a challenging RGB-D
dataset recorded in a busy operating room during live surgeries. We conduct
series of experiments to study the different part detectors in conjunction with
the various 2D or 3D pairwise constraints. Our comparisons demonstrate that 3D
PS with RGB-D part detectors significantly improves the results in a visually
challenging operating environment.
| [
{
"version": "v1",
"created": "Wed, 10 Feb 2016 17:56:47 GMT"
},
{
"version": "v2",
"created": "Mon, 22 Feb 2016 17:57:18 GMT"
},
{
"version": "v3",
"created": "Mon, 4 Jul 2016 08:56:24 GMT"
},
{
"version": "v4",
"created": "Wed, 6 Jul 2016 07:45:15 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Kadkhodamohammadi",
"Abdolrahim",
""
],
[
"Gangi",
"Afshin",
""
],
[
"de Mathelin",
"Michel",
""
],
[
"Padoy",
"Nicolas",
""
]
] | TITLE: Articulated Clinician Detection Using 3D Pictorial Structures on RGB-D
Data
ABSTRACT: Reliable human pose estimation (HPE) is essential to many clinical
applications, such as surgical workflow analysis, radiation safety monitoring
and human-robot cooperation. Proposed methods for the operating room (OR) rely
either on foreground estimation using a multi-camera system, which is a
challenge in real ORs due to color similarities and frequent illumination
changes, or on wearable sensors or markers, which are invasive and therefore
difficult to introduce in the room. Instead, we propose a novel approach based
on Pictorial Structures (PS) and on RGB-D data, which can be easily deployed in
real ORs. We extend the PS framework in two ways. First, we build robust and
discriminative part detectors using both color and depth images. We also
present a novel descriptor for depth images, called histogram of depth
differences (HDD). Second, we extend PS to 3D by proposing 3D pairwise
constraints and a new method that makes exact inference tractable. Our approach
is evaluated for pose estimation and clinician detection on a challenging RGB-D
dataset recorded in a busy operating room during live surgeries. We conduct
series of experiments to study the different part detectors in conjunction with
the various 2D or 3D pairwise constraints. Our comparisons demonstrate that 3D
PS with RGB-D part detectors significantly improves the results in a visually
challenging operating environment.
| no_new_dataset | 0.944893 |
1612.01725 | Ron Slossberg | Ron Slossberg, Aaron Wetzler and Ron Kimmel | Deep Stereo Matching with Dense CRF Priors | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Stereo reconstruction from rectified images has recently been revisited
within the context of deep learning. Using a deep Convolutional Neural Network
to obtain patch-wise matching cost volumes has resulted in state of the art
stereo reconstruction on classic datasets like Middlebury and Kitti. By
introducing this cost into a classical stereo pipeline, the final results are
improved dramatically over non-learning based cost models. However these
pipelines typically include hand engineered post processing steps to
effectively regularize and clean the result. Here, we show that it is possible
to take a more holistic approach by training a fully end-to-end network which
directly includes regularization in the form of a densely connected Conditional
Random Field (CRF) that acts as a prior on inter-pixel interactions. We
demonstrate that our approach on both synthetic and real world datasets
outperforms an alternative end-to-end network and compares favorably to more
hand engineered approaches.
| [
{
"version": "v1",
"created": "Tue, 6 Dec 2016 09:51:21 GMT"
},
{
"version": "v2",
"created": "Tue, 24 Jan 2017 20:08:28 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Slossberg",
"Ron",
""
],
[
"Wetzler",
"Aaron",
""
],
[
"Kimmel",
"Ron",
""
]
] | TITLE: Deep Stereo Matching with Dense CRF Priors
ABSTRACT: Stereo reconstruction from rectified images has recently been revisited
within the context of deep learning. Using a deep Convolutional Neural Network
to obtain patch-wise matching cost volumes has resulted in state of the art
stereo reconstruction on classic datasets like Middlebury and Kitti. By
introducing this cost into a classical stereo pipeline, the final results are
improved dramatically over non-learning based cost models. However these
pipelines typically include hand engineered post processing steps to
effectively regularize and clean the result. Here, we show that it is possible
to take a more holistic approach by training a fully end-to-end network which
directly includes regularization in the form of a densely connected Conditional
Random Field (CRF) that acts as a prior on inter-pixel interactions. We
demonstrate that our approach on both synthetic and real world datasets
outperforms an alternative end-to-end network and compares favorably to more
hand engineered approaches.
| no_new_dataset | 0.949482 |
1701.07114 | Nayyar Zaidi | Nayyar A. Zaidi, Yang Du, Geoffrey I. Webb | On the Effectiveness of Discretizing Quantitative Attributes in Linear
Classifiers | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning algorithms that learn linear models often have high representation
bias on real-world problems. In this paper, we show that this representation
bias can be greatly reduced by discretization. Discretization is a common
procedure in machine learning that is used to convert a quantitative attribute
into a qualitative one. It is often motivated by the limitation of some
learners to qualitative data. Discretization loses information, as fewer
distinctions between instances are possible using discretized data relative to
undiscretized data. In consequence, where discretization is not essential, it
might appear desirable to avoid it. However, it has been shown that
discretization often substantially reduces the error of the linear generative
Bayesian classifier naive Bayes. This motivates a systematic study of the
effectiveness of discretizing quantitative attributes for other linear
classifiers. In this work, we study the effect of discretization on the
performance of linear classifiers optimizing three distinct discriminative
objective functions --- logistic regression (optimizing negative
log-likelihood), support vector classifiers (optimizing hinge loss) and a
zero-hidden layer artificial neural network (optimizing mean-square-error). We
show that discretization can greatly increase the accuracy of these linear
discriminative learners by reducing their representation bias, especially on
big datasets. We substantiate our claims with an empirical study on $42$
benchmark datasets.
| [
{
"version": "v1",
"created": "Tue, 24 Jan 2017 23:57:32 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Zaidi",
"Nayyar A.",
""
],
[
"Du",
"Yang",
""
],
[
"Webb",
"Geoffrey I.",
""
]
] | TITLE: On the Effectiveness of Discretizing Quantitative Attributes in Linear
Classifiers
ABSTRACT: Learning algorithms that learn linear models often have high representation
bias on real-world problems. In this paper, we show that this representation
bias can be greatly reduced by discretization. Discretization is a common
procedure in machine learning that is used to convert a quantitative attribute
into a qualitative one. It is often motivated by the limitation of some
learners to qualitative data. Discretization loses information, as fewer
distinctions between instances are possible using discretized data relative to
undiscretized data. In consequence, where discretization is not essential, it
might appear desirable to avoid it. However, it has been shown that
discretization often substantially reduces the error of the linear generative
Bayesian classifier naive Bayes. This motivates a systematic study of the
effectiveness of discretizing quantitative attributes for other linear
classifiers. In this work, we study the effect of discretization on the
performance of linear classifiers optimizing three distinct discriminative
objective functions --- logistic regression (optimizing negative
log-likelihood), support vector classifiers (optimizing hinge loss) and a
zero-hidden layer artificial neural network (optimizing mean-square-error). We
show that discretization can greatly increase the accuracy of these linear
discriminative learners by reducing their representation bias, especially on
big datasets. We substantiate our claims with an empirical study on $42$
benchmark datasets.
| no_new_dataset | 0.943504 |
1701.07194 | Dacheng Tao | Shan You, Chang Xu, Yunhe Wang, Chao Xu, Dacheng Tao | Privileged Multi-label Learning | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents privileged multi-label learning (PrML) to explore and
exploit the relationship between labels in multi-label learning problems. We
suggest that for each individual label, it cannot only be implicitly connected
with other labels via the low-rank constraint over label predictors, but also
its performance on examples can receive the explicit comments from other labels
together acting as an \emph{Oracle teacher}. We generate privileged label
feature for each example and its individual label, and then integrate it into
the framework of low-rank based multi-label learning. The proposed algorithm
can therefore comprehensively explore and exploit label relationships by
inheriting all the merits of privileged information and low-rank constraints.
We show that PrML can be efficiently solved by dual coordinate descent
algorithm using iterative optimization strategy with cheap updates. Experiments
on benchmark datasets show that through privileged label features, the
performance can be significantly improved and PrML is superior to several
competing methods in most cases.
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 07:43:13 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"You",
"Shan",
""
],
[
"Xu",
"Chang",
""
],
[
"Wang",
"Yunhe",
""
],
[
"Xu",
"Chao",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: Privileged Multi-label Learning
ABSTRACT: This paper presents privileged multi-label learning (PrML) to explore and
exploit the relationship between labels in multi-label learning problems. We
suggest that for each individual label, it cannot only be implicitly connected
with other labels via the low-rank constraint over label predictors, but also
its performance on examples can receive the explicit comments from other labels
together acting as an \emph{Oracle teacher}. We generate privileged label
feature for each example and its individual label, and then integrate it into
the framework of low-rank based multi-label learning. The proposed algorithm
can therefore comprehensively explore and exploit label relationships by
inheriting all the merits of privileged information and low-rank constraints.
We show that PrML can be efficiently solved by dual coordinate descent
algorithm using iterative optimization strategy with cheap updates. Experiments
on benchmark datasets show that through privileged label features, the
performance can be significantly improved and PrML is superior to several
competing methods in most cases.
| no_new_dataset | 0.946448 |
1701.07221 | Polina Rozenshtein | Aristides Gionis, Polina Rozenshtein, Nikolaj Tatti and Evimaria Terzi | Community-aware network sparsification | null | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Network sparsification aims to reduce the number of edges of a network while
maintaining its structural properties; such properties include shortest paths,
cuts, spectral measures, or network modularity. Sparsification has multiple
applications, such as, speeding up graph-mining algorithms, graph
visualization, as well as identifying the important network edges. In this
paper we consider a novel formulation of the network-sparsification problem. In
addition to the network, we also consider as input a set of communities. The
goal is to sparsify the network so as to preserve the network structure with
respect to the given communities. We introduce two variants of the
community-aware sparsification problem, leading to sparsifiers that satisfy
different connectedness community properties. From the technical point of view,
we prove hardness results and devise effective approximation algorithms. Our
experimental results on a large collection of datasets demonstrate the
effectiveness of our algorithms.
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 09:32:15 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Gionis",
"Aristides",
""
],
[
"Rozenshtein",
"Polina",
""
],
[
"Tatti",
"Nikolaj",
""
],
[
"Terzi",
"Evimaria",
""
]
] | TITLE: Community-aware network sparsification
ABSTRACT: Network sparsification aims to reduce the number of edges of a network while
maintaining its structural properties; such properties include shortest paths,
cuts, spectral measures, or network modularity. Sparsification has multiple
applications, such as, speeding up graph-mining algorithms, graph
visualization, as well as identifying the important network edges. In this
paper we consider a novel formulation of the network-sparsification problem. In
addition to the network, we also consider as input a set of communities. The
goal is to sparsify the network so as to preserve the network structure with
respect to the given communities. We introduce two variants of the
community-aware sparsification problem, leading to sparsifiers that satisfy
different connectedness community properties. From the technical point of view,
we prove hardness results and devise effective approximation algorithms. Our
experimental results on a large collection of datasets demonstrate the
effectiveness of our algorithms.
| no_new_dataset | 0.9434 |
1701.07266 | Kfir Levy Yehuda | Oren Anava, Kfir Y. Levy | k*-Nearest Neighbors: From Global to Local | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The weighted k-nearest neighbors algorithm is one of the most fundamental
non-parametric methods in pattern recognition and machine learning. The
question of setting the optimal number of neighbors as well as the optimal
weights has received much attention throughout the years, nevertheless this
problem seems to have remained unsettled. In this paper we offer a simple
approach to locally weighted regression/classification, where we make the
bias-variance tradeoff explicit. Our formulation enables us to phrase a notion
of optimal weights, and to efficiently find these weights as well as the
optimal number of neighbors efficiently and adaptively, for each data point
whose value we wish to estimate. The applicability of our approach is
demonstrated on several datasets, showing superior performance over standard
locally weighted methods.
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 11:18:18 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Anava",
"Oren",
""
],
[
"Levy",
"Kfir Y.",
""
]
] | TITLE: k*-Nearest Neighbors: From Global to Local
ABSTRACT: The weighted k-nearest neighbors algorithm is one of the most fundamental
non-parametric methods in pattern recognition and machine learning. The
question of setting the optimal number of neighbors as well as the optimal
weights has received much attention throughout the years, nevertheless this
problem seems to have remained unsettled. In this paper we offer a simple
approach to locally weighted regression/classification, where we make the
bias-variance tradeoff explicit. Our formulation enables us to phrase a notion
of optimal weights, and to efficiently find these weights as well as the
optimal number of neighbors efficiently and adaptively, for each data point
whose value we wish to estimate. The applicability of our approach is
demonstrated on several datasets, showing superior performance over standard
locally weighted methods.
| no_new_dataset | 0.950365 |
1701.07354 | David Villacis | David Villacis, Santeri Kaupinm\"aki, Samuli Siltanen, Teemu Helenius | Photographic dataset: playing cards | 9 pages, 12 figures | null | null | null | cs.CV physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This is a photographic dataset collected for testing image processing
algorithms. The idea is to have images that can exploit the properties of total
variation, therefore a set of playing cards was distributed on the scene. The
dataset is made available at www.fips.fi/photographic_dataset2.php
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 15:35:09 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Villacis",
"David",
""
],
[
"Kaupinmäki",
"Santeri",
""
],
[
"Siltanen",
"Samuli",
""
],
[
"Helenius",
"Teemu",
""
]
] | TITLE: Photographic dataset: playing cards
ABSTRACT: This is a photographic dataset collected for testing image processing
algorithms. The idea is to have images that can exploit the properties of total
variation, therefore a set of playing cards was distributed on the scene. The
dataset is made available at www.fips.fi/photographic_dataset2.php
| new_dataset | 0.960915 |
1701.07372 | Abdolrahim Kadkhodamohammadi | Abdolrahim Kadkhodamohammadi, Afshin Gangi, Michel de Mathelin,
Nicolas Padoy | A Multi-view RGB-D Approach for Human Pose Estimation in Operating Rooms | WACV 2017. Supplementary material video: https://youtu.be/L3A0BzT0FKQ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many approaches have been proposed for human pose estimation in single and
multi-view RGB images. However, some environments, such as the operating room,
are still very challenging for state-of-the-art RGB methods. In this paper, we
propose an approach for multi-view 3D human pose estimation from RGB-D images
and demonstrate the benefits of using the additional depth channel for pose
refinement beyond its use for the generation of improved features. The proposed
method permits the joint detection and estimation of the poses without knowing
a priori the number of persons present in the scene. We evaluate this approach
on a novel multi-view RGB-D dataset acquired during live surgeries and
annotated with ground truth 3D poses.
| [
{
"version": "v1",
"created": "Wed, 25 Jan 2017 16:43:41 GMT"
}
] | 2017-01-26T00:00:00 | [
[
"Kadkhodamohammadi",
"Abdolrahim",
""
],
[
"Gangi",
"Afshin",
""
],
[
"de Mathelin",
"Michel",
""
],
[
"Padoy",
"Nicolas",
""
]
] | TITLE: A Multi-view RGB-D Approach for Human Pose Estimation in Operating Rooms
ABSTRACT: Many approaches have been proposed for human pose estimation in single and
multi-view RGB images. However, some environments, such as the operating room,
are still very challenging for state-of-the-art RGB methods. In this paper, we
propose an approach for multi-view 3D human pose estimation from RGB-D images
and demonstrate the benefits of using the additional depth channel for pose
refinement beyond its use for the generation of improved features. The proposed
method permits the joint detection and estimation of the poses without knowing
a priori the number of persons present in the scene. We evaluate this approach
on a novel multi-view RGB-D dataset acquired during live surgeries and
annotated with ground truth 3D poses.
| new_dataset | 0.960212 |
1509.00504 | Vijay Gadepally | Vijay Gadepally and Jeremy Kepner | Using a Power Law Distribution to describe Big Data | 5 pages | null | 10.1109/HPEC.2015.7322459 | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The gap between data production and user ability to access, compute and
produce meaningful results calls for tools that address the challenges
associated with big data volume, velocity and variety. One of the key hurdles
is the inability to methodically remove expected or uninteresting elements from
large data sets. This difficulty often wastes valuable researcher and
computational time by expending resources on uninteresting parts of data.
Social sensors, or sensors which produce data based on human activity, such as
Wikipedia, Twitter, and Facebook have an underlying structure which can be
thought of as having a Power Law distribution. Such a distribution implies that
few nodes generate large amounts of data. In this article, we propose a
technique to take an arbitrary dataset and compute a power law distributed
background model that bases its parameters on observed statistics. This model
can be used to determine the suitability of using a power law or automatically
identify high degree nodes for filtering and can be scaled to work with big
data.
| [
{
"version": "v1",
"created": "Fri, 28 Aug 2015 22:36:32 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Gadepally",
"Vijay",
""
],
[
"Kepner",
"Jeremy",
""
]
] | TITLE: Using a Power Law Distribution to describe Big Data
ABSTRACT: The gap between data production and user ability to access, compute and
produce meaningful results calls for tools that address the challenges
associated with big data volume, velocity and variety. One of the key hurdles
is the inability to methodically remove expected or uninteresting elements from
large data sets. This difficulty often wastes valuable researcher and
computational time by expending resources on uninteresting parts of data.
Social sensors, or sensors which produce data based on human activity, such as
Wikipedia, Twitter, and Facebook have an underlying structure which can be
thought of as having a Power Law distribution. Such a distribution implies that
few nodes generate large amounts of data. In this article, we propose a
technique to take an arbitrary dataset and compute a power law distributed
background model that bases its parameters on observed statistics. This model
can be used to determine the suitability of using a power law or automatically
identify high degree nodes for filtering and can be scaled to work with big
data.
| no_new_dataset | 0.949623 |
1510.03921 | Yongjoo Park | Yongjoo Park, Michael Cafarella, and Barzan Mozafari | Visualization-Aware Sampling for Very Large Databases | null | Data Engineering (ICDE), 2016 IEEE 32nd International Conference
on. IEEE, 2016 | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Interactive visualizations are crucial in ad hoc data exploration and
analysis. However, with the growing number of massive datasets, generating
visualizations in interactive timescales is increasingly challenging. One
approach for improving the speed of the visualization tool is via data
reduction in order to reduce the computational overhead, but at a potential
cost in visualization accuracy. Common data reduction techniques, such as
uniform and stratified sampling, do not exploit the fact that the sampled
tuples will be transformed into a visualization for human consumption.
We propose a visualization-aware sampling (VAS) that guarantees high quality
visualizations with a small subset of the entire dataset. We validate our
method when applied to scatter and map plots for three common visualization
goals: regression, density estimation, and clustering. The key to our sampling
method's success is in choosing tuples which minimize a visualization-inspired
loss function. Our user study confirms that optimizing this loss function
correlates strongly with user success in using the resulting visualizations. We
also show the NP-hardness of our optimization problem and propose an efficient
approximation algorithm. Our experiments show that, compared to previous
methods, (i) using the same sample size, VAS improves user's success by up to
35% in various visualization tasks, and (ii) VAS can achieve a required
visualization quality up to 400 times faster.
| [
{
"version": "v1",
"created": "Tue, 13 Oct 2015 22:51:36 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Jan 2017 23:47:49 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Park",
"Yongjoo",
""
],
[
"Cafarella",
"Michael",
""
],
[
"Mozafari",
"Barzan",
""
]
] | TITLE: Visualization-Aware Sampling for Very Large Databases
ABSTRACT: Interactive visualizations are crucial in ad hoc data exploration and
analysis. However, with the growing number of massive datasets, generating
visualizations in interactive timescales is increasingly challenging. One
approach for improving the speed of the visualization tool is via data
reduction in order to reduce the computational overhead, but at a potential
cost in visualization accuracy. Common data reduction techniques, such as
uniform and stratified sampling, do not exploit the fact that the sampled
tuples will be transformed into a visualization for human consumption.
We propose a visualization-aware sampling (VAS) that guarantees high quality
visualizations with a small subset of the entire dataset. We validate our
method when applied to scatter and map plots for three common visualization
goals: regression, density estimation, and clustering. The key to our sampling
method's success is in choosing tuples which minimize a visualization-inspired
loss function. Our user study confirms that optimizing this loss function
correlates strongly with user success in using the resulting visualizations. We
also show the NP-hardness of our optimization problem and propose an efficient
approximation algorithm. Our experiments show that, compared to previous
methods, (i) using the same sample size, VAS improves user's success by up to
35% in various visualization tasks, and (ii) VAS can achieve a required
visualization quality up to 400 times faster.
| no_new_dataset | 0.944944 |
1604.03199 | William Gray Roncal | William Gray Roncal, Eva L Dyer, Doga G\"ursoy, Konrad Kording,
Narayanan Kasthuri | From sample to knowledge: Towards an integrated approach for
neuroscience discovery | first two authors contributed equally. 8 pages, 2 figures. v2: added
acknowledgments | null | null | null | q-bio.QM cs.SY q-bio.NC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Imaging methods used in modern neuroscience experiments are quickly producing
large amounts of data capable of providing increasing amounts of knowledge
about neuroanatomy and function. A great deal of information in these datasets
is relatively unexplored and untapped. One of the bottlenecks in knowledge
extraction is that often there is no feedback loop between the knowledge
produced (e.g., graph, density estimate, or other statistic) and the earlier
stages of the pipeline (e.g., acquisition). We thus advocate for the
development of sample-to-knowledge discovery pipelines that one can use to
optimize acquisition and processing steps with a particular end goal (i.e.,
piece of knowledge) in mind. We therefore propose that optimization takes place
not just within each processing stage but also between adjacent (and
non-adjacent) steps of the pipeline. Furthermore, we explore the existing
categories of knowledge representation and models to motivate the types of
experiments and analysis needed to achieve the ultimate goal. To illustrate
this approach, we provide an experimental paradigm to answer questions about
large-scale synaptic distributions through a multimodal approach combining
X-ray microtomography and electron microscopy.
| [
{
"version": "v1",
"created": "Tue, 12 Apr 2016 01:41:48 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Jan 2017 19:30:41 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Roncal",
"William Gray",
""
],
[
"Dyer",
"Eva L",
""
],
[
"Gürsoy",
"Doga",
""
],
[
"Kording",
"Konrad",
""
],
[
"Kasthuri",
"Narayanan",
""
]
] | TITLE: From sample to knowledge: Towards an integrated approach for
neuroscience discovery
ABSTRACT: Imaging methods used in modern neuroscience experiments are quickly producing
large amounts of data capable of providing increasing amounts of knowledge
about neuroanatomy and function. A great deal of information in these datasets
is relatively unexplored and untapped. One of the bottlenecks in knowledge
extraction is that often there is no feedback loop between the knowledge
produced (e.g., graph, density estimate, or other statistic) and the earlier
stages of the pipeline (e.g., acquisition). We thus advocate for the
development of sample-to-knowledge discovery pipelines that one can use to
optimize acquisition and processing steps with a particular end goal (i.e.,
piece of knowledge) in mind. We therefore propose that optimization takes place
not just within each processing stage but also between adjacent (and
non-adjacent) steps of the pipeline. Furthermore, we explore the existing
categories of knowledge representation and models to motivate the types of
experiments and analysis needed to achieve the ultimate goal. To illustrate
this approach, we provide an experimental paradigm to answer questions about
large-scale synaptic distributions through a multimodal approach combining
X-ray microtomography and electron microscopy.
| no_new_dataset | 0.945701 |
1607.05258 | Mohammad Havaei | Mohammad Havaei and Nicolas Guizard and Hugo Larochelle and
Pierre-Marc Jodoin | Deep learning trends for focal brain pathology segmentation in MRI | Published in Machine Learning for Health Informatics | null | 10.1007/978-3-319-50478-0_6 | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Segmentation of focal (localized) brain pathologies such as brain tumors and
brain lesions caused by multiple sclerosis and ischemic strokes are necessary
for medical diagnosis, surgical planning and disease development as well as
other applications such as tractography. Over the years, attempts have been
made to automate this process for both clinical and research reasons. In this
regard, machine learning methods have long been a focus of attention. Over the
past two years, the medical imaging field has seen a rise in the use of a
particular branch of machine learning commonly known as deep learning. In the
non-medical computer vision world, deep learning based methods have obtained
state-of-the-art results on many datasets. Recent studies in computer aided
diagnostics have shown deep learning methods (and especially convolutional
neural networks - CNN) to yield promising results. In this chapter, we provide
a survey of CNN methods applied to medical imaging with a focus on brain
pathology segmentation. In particular, we discuss their characteristic
peculiarities and their specific configuration and adjustments that are best
suited to segment medical images. We also underline the intrinsic differences
deep learning methods have with other machine learning methods.
| [
{
"version": "v1",
"created": "Mon, 18 Jul 2016 19:52:00 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Jan 2017 16:41:46 GMT"
},
{
"version": "v3",
"created": "Tue, 24 Jan 2017 02:44:48 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Havaei",
"Mohammad",
""
],
[
"Guizard",
"Nicolas",
""
],
[
"Larochelle",
"Hugo",
""
],
[
"Jodoin",
"Pierre-Marc",
""
]
] | TITLE: Deep learning trends for focal brain pathology segmentation in MRI
ABSTRACT: Segmentation of focal (localized) brain pathologies such as brain tumors and
brain lesions caused by multiple sclerosis and ischemic strokes are necessary
for medical diagnosis, surgical planning and disease development as well as
other applications such as tractography. Over the years, attempts have been
made to automate this process for both clinical and research reasons. In this
regard, machine learning methods have long been a focus of attention. Over the
past two years, the medical imaging field has seen a rise in the use of a
particular branch of machine learning commonly known as deep learning. In the
non-medical computer vision world, deep learning based methods have obtained
state-of-the-art results on many datasets. Recent studies in computer aided
diagnostics have shown deep learning methods (and especially convolutional
neural networks - CNN) to yield promising results. In this chapter, we provide
a survey of CNN methods applied to medical imaging with a focus on brain
pathology segmentation. In particular, we discuss their characteristic
peculiarities and their specific configuration and adjustments that are best
suited to segment medical images. We also underline the intrinsic differences
deep learning methods have with other machine learning methods.
| no_new_dataset | 0.946547 |
1701.06643 | Sergey Korolev | Sergey Korolev, Amir Safiullin, Mikhail Belyaev, Yulia Dodonova | Residual and Plain Convolutional Neural Networks for 3D Brain MRI
Classification | IEEE International Symposium on Biomedical Imaging 2017 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the recent years there have been a number of studies that applied deep
learning algorithms to neuroimaging data. Pipelines used in those studies
mostly require multiple processing steps for feature extraction, although
modern advancements in deep learning for image classification can provide a
powerful framework for automatic feature generation and more straightforward
analysis. In this paper, we show how similar performance can be achieved
skipping these feature extraction steps with the residual and plain 3D
convolutional neural network architectures. We demonstrate the performance of
the proposed approach for classification of Alzheimer's disease versus mild
cognitive impairment and normal controls on the Alzheimer's Disease National
Initiative (ADNI) dataset of 3D structural MRI brain scans.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 21:54:50 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Korolev",
"Sergey",
""
],
[
"Safiullin",
"Amir",
""
],
[
"Belyaev",
"Mikhail",
""
],
[
"Dodonova",
"Yulia",
""
]
] | TITLE: Residual and Plain Convolutional Neural Networks for 3D Brain MRI
Classification
ABSTRACT: In the recent years there have been a number of studies that applied deep
learning algorithms to neuroimaging data. Pipelines used in those studies
mostly require multiple processing steps for feature extraction, although
modern advancements in deep learning for image classification can provide a
powerful framework for automatic feature generation and more straightforward
analysis. In this paper, we show how similar performance can be achieved
skipping these feature extraction steps with the residual and plain 3D
convolutional neural network architectures. We demonstrate the performance of
the proposed approach for classification of Alzheimer's disease versus mild
cognitive impairment and normal controls on the Alzheimer's Disease National
Initiative (ADNI) dataset of 3D structural MRI brain scans.
| no_new_dataset | 0.949716 |
1701.06659 | Cheng-Yang Fu | Cheng-Yang Fu, Wei Liu, Ananth Ranga, Ambrish Tyagi, Alexander C. Berg | DSSD : Deconvolutional Single Shot Detector | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The main contribution of this paper is an approach for introducing additional
context into state-of-the-art general object detection. To achieve this we
first combine a state-of-the-art classifier (Residual-101[14]) with a fast
detection framework (SSD[18]). We then augment SSD+Residual-101 with
deconvolution layers to introduce additional large-scale context in object
detection and improve accuracy, especially for small objects, calling our
resulting system DSSD for deconvolutional single shot detector. While these two
contributions are easily described at a high-level, a naive implementation does
not succeed. Instead we show that carefully adding additional stages of learned
transformations, specifically a module for feed-forward connections in
deconvolution and a new output module, enables this new approach and forms a
potential way forward for further detection research. Results are shown on both
PASCAL VOC and COCO detection. Our DSSD with $513 \times 513$ input achieves
81.5% mAP on VOC2007 test, 80.0% mAP on VOC2012 test, and 33.2% mAP on COCO,
outperforming a state-of-the-art method R-FCN[3] on each dataset.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 22:33:35 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Fu",
"Cheng-Yang",
""
],
[
"Liu",
"Wei",
""
],
[
"Ranga",
"Ananth",
""
],
[
"Tyagi",
"Ambrish",
""
],
[
"Berg",
"Alexander C.",
""
]
] | TITLE: DSSD : Deconvolutional Single Shot Detector
ABSTRACT: The main contribution of this paper is an approach for introducing additional
context into state-of-the-art general object detection. To achieve this we
first combine a state-of-the-art classifier (Residual-101[14]) with a fast
detection framework (SSD[18]). We then augment SSD+Residual-101 with
deconvolution layers to introduce additional large-scale context in object
detection and improve accuracy, especially for small objects, calling our
resulting system DSSD for deconvolutional single shot detector. While these two
contributions are easily described at a high-level, a naive implementation does
not succeed. Instead we show that carefully adding additional stages of learned
transformations, specifically a module for feed-forward connections in
deconvolution and a new output module, enables this new approach and forms a
potential way forward for further detection research. Results are shown on both
PASCAL VOC and COCO detection. Our DSSD with $513 \times 513$ input achieves
81.5% mAP on VOC2007 test, 80.0% mAP on VOC2012 test, and 33.2% mAP on COCO,
outperforming a state-of-the-art method R-FCN[3] on each dataset.
| no_new_dataset | 0.947624 |
1701.06715 | Xiaohao Cai | Juheon Lee, David Coomes, Carola-Bibiane Schonlieb, Xiaohao Cai, Jan
Lellmann, Michele Dalponte, Yadvinder Malhi, Nathalie Butt, Mike Morecroft | A graph cut approach to 3D tree delineation, using integrated airborne
LiDAR and hyperspectral imagery | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recognising individual trees within remotely sensed imagery has important
applications in forest ecology and management. Several algorithms for tree
delineation have been suggested, mostly based on locating local maxima or
inverted basins in raster canopy height models (CHMs) derived from Light
Detection And Ranging (LiDAR) data or photographs. However, these algorithms
often lead to inaccurate estimates of forest stand characteristics due to the
limited information content of raster CHMs. Here we develop a 3D tree
delineation method which uses graph cut to delineate trees from the full 3D
LiDAR point cloud, and also makes use of any optical imagery available
(hyperspectral imagery in our case). First, conventional methods are used to
locate local maxima in the CHM and generate an initial map of trees. Second, a
graph is built from the LiDAR point cloud, fused with the hyperspectral data.
For computational efficiency, the feature space of hyperspectral imagery is
reduced using robust PCA. Third, a multi-class normalised cut is applied to the
graph, using the initial map of trees to constrain the number of clusters and
their locations. Finally, recursive normalised cut is used to subdivide, if
necessary, each of the clusters identified by the initial analysis. We call
this approach Multiclass Cut followed by Recursive Cut (MCRC). The
effectiveness of MCRC was tested using three datasets: i) NewFor, ii) a
coniferous forest in the Italian Alps, and iii) a deciduous woodland in the UK.
The performance of MCRC was usually superior to that of other delineation
methods, and was further improved by including high-resolution optical imagery.
Since MCRC delineates the entire LiDAR point cloud in 3D, it allows individual
crown characteristics to be measured. By making full use of the data available,
graph cut has the potential to considerably improve the accuracy of tree
delineation.
| [
{
"version": "v1",
"created": "Tue, 24 Jan 2017 02:41:30 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Lee",
"Juheon",
""
],
[
"Coomes",
"David",
""
],
[
"Schonlieb",
"Carola-Bibiane",
""
],
[
"Cai",
"Xiaohao",
""
],
[
"Lellmann",
"Jan",
""
],
[
"Dalponte",
"Michele",
""
],
[
"Malhi",
"Yadvinder",
""
],
[
"Butt",
"Nathalie",
""
],
[
"Morecroft",
"Mike",
""
]
] | TITLE: A graph cut approach to 3D tree delineation, using integrated airborne
LiDAR and hyperspectral imagery
ABSTRACT: Recognising individual trees within remotely sensed imagery has important
applications in forest ecology and management. Several algorithms for tree
delineation have been suggested, mostly based on locating local maxima or
inverted basins in raster canopy height models (CHMs) derived from Light
Detection And Ranging (LiDAR) data or photographs. However, these algorithms
often lead to inaccurate estimates of forest stand characteristics due to the
limited information content of raster CHMs. Here we develop a 3D tree
delineation method which uses graph cut to delineate trees from the full 3D
LiDAR point cloud, and also makes use of any optical imagery available
(hyperspectral imagery in our case). First, conventional methods are used to
locate local maxima in the CHM and generate an initial map of trees. Second, a
graph is built from the LiDAR point cloud, fused with the hyperspectral data.
For computational efficiency, the feature space of hyperspectral imagery is
reduced using robust PCA. Third, a multi-class normalised cut is applied to the
graph, using the initial map of trees to constrain the number of clusters and
their locations. Finally, recursive normalised cut is used to subdivide, if
necessary, each of the clusters identified by the initial analysis. We call
this approach Multiclass Cut followed by Recursive Cut (MCRC). The
effectiveness of MCRC was tested using three datasets: i) NewFor, ii) a
coniferous forest in the Italian Alps, and iii) a deciduous woodland in the UK.
The performance of MCRC was usually superior to that of other delineation
methods, and was further improved by including high-resolution optical imagery.
Since MCRC delineates the entire LiDAR point cloud in 3D, it allows individual
crown characteristics to be measured. By making full use of the data available,
graph cut has the potential to considerably improve the accuracy of tree
delineation.
| no_new_dataset | 0.950549 |
1701.06751 | Qiongkai Xu | Qiongkai Xu, Qing Wang, Chenchen Xu and Lizhen Qu | Collective Vertex Classification Using Recursive Neural Network | 7 pages, 5 figures | null | null | null | cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Collective classification of vertices is a task of assigning categories to
each vertex in a graph based on both vertex attributes and link structure.
Nevertheless, some existing approaches do not use the features of neighbouring
vertices properly, due to the noise introduced by these features. In this
paper, we propose a graph-based recursive neural network framework for
collective vertex classification. In this framework, we generate hidden
representations from both attributes of vertices and representations of
neighbouring vertices via recursive neural networks. Under this framework, we
explore two types of recursive neural units, naive recursive neural unit and
long short-term memory unit. We have conducted experiments on four real-world
network datasets. The experimental results show that our frame- work with long
short-term memory model achieves better results and outperforms several
competitive baseline methods.
| [
{
"version": "v1",
"created": "Tue, 24 Jan 2017 07:07:15 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Xu",
"Qiongkai",
""
],
[
"Wang",
"Qing",
""
],
[
"Xu",
"Chenchen",
""
],
[
"Qu",
"Lizhen",
""
]
] | TITLE: Collective Vertex Classification Using Recursive Neural Network
ABSTRACT: Collective classification of vertices is a task of assigning categories to
each vertex in a graph based on both vertex attributes and link structure.
Nevertheless, some existing approaches do not use the features of neighbouring
vertices properly, due to the noise introduced by these features. In this
paper, we propose a graph-based recursive neural network framework for
collective vertex classification. In this framework, we generate hidden
representations from both attributes of vertices and representations of
neighbouring vertices via recursive neural networks. Under this framework, we
explore two types of recursive neural units, naive recursive neural unit and
long short-term memory unit. We have conducted experiments on four real-world
network datasets. The experimental results show that our frame- work with long
short-term memory model achieves better results and outperforms several
competitive baseline methods.
| no_new_dataset | 0.951729 |
1701.06861 | Panagiotis Papapetrou | Hend Kareem, Lars Asker, and Panagiotis Papapetrou | Detecting Hierarchical Ties Using Link-Analysis Ranking at Different
Levels of Time Granularity | null | null | null | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Social networks contain implicit knowledge that can be used to infer
hierarchical relations that are not explicitly present in the available data.
Interaction patterns are typically affected by users' social relations. We
present an approach to inferring such information that applies a link-analysis
ranking algorithm at different levels of time granularity. In addition, a
voting scheme is employed for obtaining the hierarchical relations. The
approach is evaluated on two datasets: the Enron email data set, where the goal
is to infer manager-subordinate relationships, and the Co-author data set,
where the goal is to infer PhD advisor-advisee relations. The experimental
results indicate that the proposed approach outperforms more traditional
approaches to inferring hierarchical relations from social networks.
| [
{
"version": "v1",
"created": "Tue, 24 Jan 2017 13:23:40 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Kareem",
"Hend",
""
],
[
"Asker",
"Lars",
""
],
[
"Papapetrou",
"Panagiotis",
""
]
] | TITLE: Detecting Hierarchical Ties Using Link-Analysis Ranking at Different
Levels of Time Granularity
ABSTRACT: Social networks contain implicit knowledge that can be used to infer
hierarchical relations that are not explicitly present in the available data.
Interaction patterns are typically affected by users' social relations. We
present an approach to inferring such information that applies a link-analysis
ranking algorithm at different levels of time granularity. In addition, a
voting scheme is employed for obtaining the hierarchical relations. The
approach is evaluated on two datasets: the Enron email data set, where the goal
is to infer manager-subordinate relationships, and the Co-author data set,
where the goal is to infer PhD advisor-advisee relations. The experimental
results indicate that the proposed approach outperforms more traditional
approaches to inferring hierarchical relations from social networks.
| no_new_dataset | 0.948298 |
1701.06944 | Michael Ying Yang | Michael Ying Yang, Hanno Ackermann, Weiyao Lin, Sitong Feng, Bodo
Rosenhahn | Motion Segmentation via Global and Local Sparse Subspace Optimization | 11 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we propose a new framework for segmenting feature-based moving
objects under affine subspace model. Since the feature trajectories in practice
are high-dimensional and contain a lot of noise, we firstly apply the sparse
PCA to represent the original trajectories with a low-dimensional global
subspace, which consists of the orthogonal sparse principal vectors.
Subsequently, the local subspace separation will be achieved via automatically
searching the sparse representation of the nearest neighbors for each projected
data. In order to refine the local subspace estimation result and deal with the
missing data problem, we propose an error estimation to encourage the projected
data that span a same local subspace to be clustered together. In the end, the
segmentation of different motions is achieved through the spectral clustering
on an affinity matrix, which is constructed with both the error estimation and
sparse neighbors optimization. We test our method extensively and compare it
with state-of-the-art methods on the Hopkins 155 dataset and Freiburg-Berkeley
Motion Segmentation dataset. The results show that our method is comparable
with the other motion segmentation methods, and in many cases exceed them in
terms of precision and computation time.
| [
{
"version": "v1",
"created": "Tue, 24 Jan 2017 15:49:53 GMT"
}
] | 2017-01-25T00:00:00 | [
[
"Yang",
"Michael Ying",
""
],
[
"Ackermann",
"Hanno",
""
],
[
"Lin",
"Weiyao",
""
],
[
"Feng",
"Sitong",
""
],
[
"Rosenhahn",
"Bodo",
""
]
] | TITLE: Motion Segmentation via Global and Local Sparse Subspace Optimization
ABSTRACT: In this paper, we propose a new framework for segmenting feature-based moving
objects under affine subspace model. Since the feature trajectories in practice
are high-dimensional and contain a lot of noise, we firstly apply the sparse
PCA to represent the original trajectories with a low-dimensional global
subspace, which consists of the orthogonal sparse principal vectors.
Subsequently, the local subspace separation will be achieved via automatically
searching the sparse representation of the nearest neighbors for each projected
data. In order to refine the local subspace estimation result and deal with the
missing data problem, we propose an error estimation to encourage the projected
data that span a same local subspace to be clustered together. In the end, the
segmentation of different motions is achieved through the spectral clustering
on an affinity matrix, which is constructed with both the error estimation and
sparse neighbors optimization. We test our method extensively and compare it
with state-of-the-art methods on the Hopkins 155 dataset and Freiburg-Berkeley
Motion Segmentation dataset. The results show that our method is comparable
with the other motion segmentation methods, and in many cases exceed them in
terms of precision and computation time.
| no_new_dataset | 0.948585 |
1507.04576 | Maedeh Aghaei | Maedeh Aghaei and Mariella Dimiccoli and Petia Radeva | Multi-Face Tracking by Extended Bag-of-Tracklets in Egocentric Videos | 27 pages, 18 figures, submitted to computer vision and image
understanding journal | null | 10.1016/j.cviu.2016.02.013 | YCVIU2393 | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Wearable cameras offer a hands-free way to record egocentric images of daily
experiences, where social events are of special interest. The first step
towards detection of social events is to track the appearance of multiple
persons involved in it. In this paper, we propose a novel method to find
correspondences of multiple faces in low temporal resolution egocentric videos
acquired through a wearable camera. This kind of photo-stream imposes
additional challenges to the multi-tracking problem with respect to
conventional videos. Due to the free motion of the camera and to its low
temporal resolution, abrupt changes in the field of view, in illumination
condition and in the target location are highly frequent. To overcome such
difficulties, we propose a multi-face tracking method that generates a set of
tracklets through finding correspondences along the whole sequence for each
detected face and takes advantage of the tracklets redundancy to deal with
unreliable ones. Similar tracklets are grouped into the so called extended
bag-of-tracklets (eBoT), which is aimed to correspond to a specific person.
Finally, a prototype tracklet is extracted for each eBoT, where the occurred
occlusions are estimated by relying on a new measure of confidence. We
validated our approach over an extensive dataset of egocentric photo-streams
and compared it to state of the art methods, demonstrating its effectiveness
and robustness.
| [
{
"version": "v1",
"created": "Thu, 16 Jul 2015 13:51:47 GMT"
},
{
"version": "v2",
"created": "Wed, 13 Jan 2016 12:26:09 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Aghaei",
"Maedeh",
""
],
[
"Dimiccoli",
"Mariella",
""
],
[
"Radeva",
"Petia",
""
]
] | TITLE: Multi-Face Tracking by Extended Bag-of-Tracklets in Egocentric Videos
ABSTRACT: Wearable cameras offer a hands-free way to record egocentric images of daily
experiences, where social events are of special interest. The first step
towards detection of social events is to track the appearance of multiple
persons involved in it. In this paper, we propose a novel method to find
correspondences of multiple faces in low temporal resolution egocentric videos
acquired through a wearable camera. This kind of photo-stream imposes
additional challenges to the multi-tracking problem with respect to
conventional videos. Due to the free motion of the camera and to its low
temporal resolution, abrupt changes in the field of view, in illumination
condition and in the target location are highly frequent. To overcome such
difficulties, we propose a multi-face tracking method that generates a set of
tracklets through finding correspondences along the whole sequence for each
detected face and takes advantage of the tracklets redundancy to deal with
unreliable ones. Similar tracklets are grouped into the so called extended
bag-of-tracklets (eBoT), which is aimed to correspond to a specific person.
Finally, a prototype tracklet is extracted for each eBoT, where the occurred
occlusions are estimated by relying on a new measure of confidence. We
validated our approach over an extensive dataset of egocentric photo-streams
and compared it to state of the art methods, demonstrating its effectiveness
and robustness.
| no_new_dataset | 0.923247 |
1605.05590 | Matteo Ceccarello | Matteo Ceccarello, Andrea Pietracaprina, Geppino Pucci, Eli Upfal | MapReduce and Streaming Algorithms for Diversity Maximization in Metric
Spaces of Bounded Doubling Dimension | Extended version of
http://www.vldb.org/pvldb/vol10/p469-ceccarello.pdf, PVLDB Volume 10, No. 5,
January 2017 | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given a dataset of points in a metric space and an integer $k$, a diversity
maximization problem requires determining a subset of $k$ points maximizing
some diversity objective measure, e.g., the minimum or the average distance
between two points in the subset. Diversity maximization is computationally
hard, hence only approximate solutions can be hoped for. Although its
applications are mainly in massive data analysis, most of the past research on
diversity maximization focused on the sequential setting. In this work we
present space and pass/round-efficient diversity maximization algorithms for
the Streaming and MapReduce models and analyze their approximation guarantees
for the relevant class of metric spaces of bounded doubling dimension. Like
other approaches in the literature, our algorithms rely on the determination of
high-quality core-sets, i.e., (much) smaller subsets of the input which contain
good approximations to the optimal solution for the whole input. For a variety
of diversity objective functions, our algorithms attain an
$(\alpha+\epsilon)$-approximation ratio, for any constant $\epsilon>0$, where
$\alpha$ is the best approximation ratio achieved by a polynomial-time,
linear-space sequential algorithm for the same diversity objective. This
improves substantially over the approximation ratios attainable in Streaming
and MapReduce by state-of-the-art algorithms for general metric spaces. We
provide extensive experimental evidence of the effectiveness of our algorithms
on both real world and synthetic datasets, scaling up to over a billion points.
| [
{
"version": "v1",
"created": "Wed, 18 May 2016 14:11:31 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Jun 2016 12:55:52 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Oct 2016 13:04:51 GMT"
},
{
"version": "v4",
"created": "Mon, 23 Jan 2017 16:10:19 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Ceccarello",
"Matteo",
""
],
[
"Pietracaprina",
"Andrea",
""
],
[
"Pucci",
"Geppino",
""
],
[
"Upfal",
"Eli",
""
]
] | TITLE: MapReduce and Streaming Algorithms for Diversity Maximization in Metric
Spaces of Bounded Doubling Dimension
ABSTRACT: Given a dataset of points in a metric space and an integer $k$, a diversity
maximization problem requires determining a subset of $k$ points maximizing
some diversity objective measure, e.g., the minimum or the average distance
between two points in the subset. Diversity maximization is computationally
hard, hence only approximate solutions can be hoped for. Although its
applications are mainly in massive data analysis, most of the past research on
diversity maximization focused on the sequential setting. In this work we
present space and pass/round-efficient diversity maximization algorithms for
the Streaming and MapReduce models and analyze their approximation guarantees
for the relevant class of metric spaces of bounded doubling dimension. Like
other approaches in the literature, our algorithms rely on the determination of
high-quality core-sets, i.e., (much) smaller subsets of the input which contain
good approximations to the optimal solution for the whole input. For a variety
of diversity objective functions, our algorithms attain an
$(\alpha+\epsilon)$-approximation ratio, for any constant $\epsilon>0$, where
$\alpha$ is the best approximation ratio achieved by a polynomial-time,
linear-space sequential algorithm for the same diversity objective. This
improves substantially over the approximation ratios attainable in Streaming
and MapReduce by state-of-the-art algorithms for general metric spaces. We
provide extensive experimental evidence of the effectiveness of our algorithms
on both real world and synthetic datasets, scaling up to over a billion points.
| no_new_dataset | 0.947962 |
1605.06276 | Alexander Gorban | A.N. Gorban, E.M. Mirkes, A. Zinovyev | Piece-wise quadratic approximations of arbitrary error functions for
fast and robust machine learning | Edited and extended version with algortihms of regularized regression | Neural Networks, Volume 84, December 2016, 28-38 | 10.1016/j.neunet.2016.08.007 | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Most of machine learning approaches have stemmed from the application of
minimizing the mean squared distance principle, based on the computationally
efficient quadratic optimization methods. However, when faced with
high-dimensional and noisy data, the quadratic error functionals demonstrated
many weaknesses including high sensitivity to contaminating factors and
dimensionality curse. Therefore, a lot of recent applications in machine
learning exploited properties of non-quadratic error functionals based on $L_1$
norm or even sub-linear potentials corresponding to quasinorms $L_p$ ($0<p<1$).
The back side of these approaches is increase in computational cost for
optimization. Till so far, no approaches have been suggested to deal with {\it
arbitrary} error functionals, in a flexible and computationally efficient
framework. In this paper, we develop a theory and basic universal data
approximation algorithms ($k$-means, principal components, principal manifolds
and graphs, regularized and sparse regression), based on piece-wise quadratic
error potentials of subquadratic growth (PQSQ potentials). We develop a new and
universal framework to minimize {\it arbitrary sub-quadratic error potentials}
using an algorithm with guaranteed fast convergence to the local or global
error minimum. The theory of PQSQ potentials is based on the notion of the cone
of minorant functions, and represents a natural approximation formalism based
on the application of min-plus algebra. The approach can be applied in most of
existing machine learning methods, including methods of data approximation and
regularized and sparse regression, leading to the improvement in the
computational cost/accuracy trade-off. We demonstrate that on synthetic and
real-life datasets PQSQ-based machine learning methods achieve orders of
magnitude faster computational performance than the corresponding
state-of-the-art methods.
| [
{
"version": "v1",
"created": "Fri, 20 May 2016 10:25:47 GMT"
},
{
"version": "v2",
"created": "Sun, 21 Aug 2016 12:44:25 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Gorban",
"A. N.",
""
],
[
"Mirkes",
"E. M.",
""
],
[
"Zinovyev",
"A.",
""
]
] | TITLE: Piece-wise quadratic approximations of arbitrary error functions for
fast and robust machine learning
ABSTRACT: Most of machine learning approaches have stemmed from the application of
minimizing the mean squared distance principle, based on the computationally
efficient quadratic optimization methods. However, when faced with
high-dimensional and noisy data, the quadratic error functionals demonstrated
many weaknesses including high sensitivity to contaminating factors and
dimensionality curse. Therefore, a lot of recent applications in machine
learning exploited properties of non-quadratic error functionals based on $L_1$
norm or even sub-linear potentials corresponding to quasinorms $L_p$ ($0<p<1$).
The back side of these approaches is increase in computational cost for
optimization. Till so far, no approaches have been suggested to deal with {\it
arbitrary} error functionals, in a flexible and computationally efficient
framework. In this paper, we develop a theory and basic universal data
approximation algorithms ($k$-means, principal components, principal manifolds
and graphs, regularized and sparse regression), based on piece-wise quadratic
error potentials of subquadratic growth (PQSQ potentials). We develop a new and
universal framework to minimize {\it arbitrary sub-quadratic error potentials}
using an algorithm with guaranteed fast convergence to the local or global
error minimum. The theory of PQSQ potentials is based on the notion of the cone
of minorant functions, and represents a natural approximation formalism based
on the application of min-plus algebra. The approach can be applied in most of
existing machine learning methods, including methods of data approximation and
regularized and sparse regression, leading to the improvement in the
computational cost/accuracy trade-off. We demonstrate that on synthetic and
real-life datasets PQSQ-based machine learning methods achieve orders of
magnitude faster computational performance than the corresponding
state-of-the-art methods.
| no_new_dataset | 0.946498 |
1607.05836 | Jiaping Zhao | Jiaping Zhao and Laurent Itti | Improved Deep Learning of Object Category using Pose Information | 10 pages, accepted by WACV 2017 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite significant recent progress, the best available computer vision
algorithms still lag far behind human capabilities, even for recognizing
individual discrete objects under various poses, illuminations, and
backgrounds. Here we present a new approach to using object pose information to
improve deep network learning. While existing large-scale datasets, e.g.
ImageNet, do not have pose information, we leverage the newly published
turntable dataset, iLab-20M, which has ~22M images of 704 object instances shot
under different lightings, camera viewpoints and turntable rotations, to do
more controlled object recognition experiments. We introduce a new
convolutional neural network architecture, what/where CNN (2W-CNN), built on a
linear-chain feedforward CNN (e.g., AlexNet), augmented by hierarchical layers
regularized by object poses. Pose information is only used as feedback signal
during training, in addition to category information; during test, the
feedforward network only predicts category. To validate the approach, we train
both 2W-CNN and AlexNet using a fraction of the dataset, and 2W-CNN achieves 6%
performance improvement in category prediction. We show mathematically that
2W-CNN has inherent advantages over AlexNet under the stochastic gradient
descent (SGD) optimization procedure. Further more, we fine-tune object
recognition on ImageNet by using the pretrained 2W-CNN and AlexNet features on
iLab-20M, results show that significant improvements have been achieved,
compared with training AlexNet from scratch. Moreover, fine-tuning 2W-CNN
features performs even better than fine-tuning the pretrained AlexNet features.
These results show pretrained features on iLab- 20M generalizes well to natural
image datasets, and 2WCNN learns even better features for object recognition
than AlexNet.
| [
{
"version": "v1",
"created": "Wed, 20 Jul 2016 07:11:08 GMT"
},
{
"version": "v2",
"created": "Wed, 31 Aug 2016 19:07:10 GMT"
},
{
"version": "v3",
"created": "Sun, 22 Jan 2017 23:53:15 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Zhao",
"Jiaping",
""
],
[
"Itti",
"Laurent",
""
]
] | TITLE: Improved Deep Learning of Object Category using Pose Information
ABSTRACT: Despite significant recent progress, the best available computer vision
algorithms still lag far behind human capabilities, even for recognizing
individual discrete objects under various poses, illuminations, and
backgrounds. Here we present a new approach to using object pose information to
improve deep network learning. While existing large-scale datasets, e.g.
ImageNet, do not have pose information, we leverage the newly published
turntable dataset, iLab-20M, which has ~22M images of 704 object instances shot
under different lightings, camera viewpoints and turntable rotations, to do
more controlled object recognition experiments. We introduce a new
convolutional neural network architecture, what/where CNN (2W-CNN), built on a
linear-chain feedforward CNN (e.g., AlexNet), augmented by hierarchical layers
regularized by object poses. Pose information is only used as feedback signal
during training, in addition to category information; during test, the
feedforward network only predicts category. To validate the approach, we train
both 2W-CNN and AlexNet using a fraction of the dataset, and 2W-CNN achieves 6%
performance improvement in category prediction. We show mathematically that
2W-CNN has inherent advantages over AlexNet under the stochastic gradient
descent (SGD) optimization procedure. Further more, we fine-tune object
recognition on ImageNet by using the pretrained 2W-CNN and AlexNet features on
iLab-20M, results show that significant improvements have been achieved,
compared with training AlexNet from scratch. Moreover, fine-tuning 2W-CNN
features performs even better than fine-tuning the pretrained AlexNet features.
These results show pretrained features on iLab- 20M generalizes well to natural
image datasets, and 2WCNN learns even better features for object recognition
than AlexNet.
| no_new_dataset | 0.946597 |
1607.05851 | Jiaping Zhao | Jiaping Zhao, Chin-kai Chang and Laurent Itti | Learning to Recognize Objects by Retaining other Factors of Variation | 9 pages, accepted by WACV 2017 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Natural images are generated under many factors, including shape, pose,
illumination etc. Most existing ConvNets formulate object recognition from
natural images as a single task classification problem, and attempt to learn
features useful for object categories, but invariant to other factors of
variation as much as possible. These architectures do not explicitly learn
other factors, like pose and lighting, instead, they usually discard them by
pooling and normalization. In this work, we take the opposite approach: we
train ConvNets for object recognition by retaining other factors (pose in our
case) and learn them jointly with object category. We design a new multi-task
leaning (MTL) ConvNet, named disentangling CNN (disCNN), which explicitly
enforces the disentangled representations of object identity and pose, and is
trained to predict object categories and pose transformations. We show that
disCNN achieves significantly better object recognition accuracies than AlexNet
trained solely to predict object categories on the iLab-20M dataset, which is a
large scale turntable dataset with detailed object pose and lighting
information. We further show that the pretrained disCNN/AlexNet features on
iLab- 20M generalize to object recognition on both Washington RGB-D and
ImageNet datasets, and the pretrained disCNN features are significantly better
than the pretrained AlexNet features for fine-tuning object recognition on the
ImageNet dataset.
| [
{
"version": "v1",
"created": "Wed, 20 Jul 2016 07:58:57 GMT"
},
{
"version": "v2",
"created": "Wed, 31 Aug 2016 19:05:35 GMT"
},
{
"version": "v3",
"created": "Sun, 22 Jan 2017 23:56:42 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Zhao",
"Jiaping",
""
],
[
"Chang",
"Chin-kai",
""
],
[
"Itti",
"Laurent",
""
]
] | TITLE: Learning to Recognize Objects by Retaining other Factors of Variation
ABSTRACT: Natural images are generated under many factors, including shape, pose,
illumination etc. Most existing ConvNets formulate object recognition from
natural images as a single task classification problem, and attempt to learn
features useful for object categories, but invariant to other factors of
variation as much as possible. These architectures do not explicitly learn
other factors, like pose and lighting, instead, they usually discard them by
pooling and normalization. In this work, we take the opposite approach: we
train ConvNets for object recognition by retaining other factors (pose in our
case) and learn them jointly with object category. We design a new multi-task
leaning (MTL) ConvNet, named disentangling CNN (disCNN), which explicitly
enforces the disentangled representations of object identity and pose, and is
trained to predict object categories and pose transformations. We show that
disCNN achieves significantly better object recognition accuracies than AlexNet
trained solely to predict object categories on the iLab-20M dataset, which is a
large scale turntable dataset with detailed object pose and lighting
information. We further show that the pretrained disCNN/AlexNet features on
iLab- 20M generalize to object recognition on both Washington RGB-D and
ImageNet datasets, and the pretrained disCNN features are significantly better
than the pretrained AlexNet features for fine-tuning object recognition on the
ImageNet dataset.
| no_new_dataset | 0.948822 |
1609.00085 | Rajasekar Venkatesan | Rajasekar Venkatesan, Meng Joo Er | A Novel Progressive Learning Technique for Multi-class Classification | 23 pages, 13 tables, 11 figures | null | null | null | cs.LG cs.AI cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a progressive learning technique for multi-class
classification is proposed. This newly developed learning technique is
independent of the number of class constraints and it can learn new classes
while still retaining the knowledge of previous classes. Whenever a new class
(non-native to the knowledge learnt thus far) is encountered, the neural
network structure gets remodeled automatically by facilitating new neurons and
interconnections, and the parameters are calculated in such a way that it
retains the knowledge learnt thus far. This technique is suitable for
real-world applications where the number of classes is often unknown and online
learning from real-time data is required. The consistency and the complexity of
the progressive learning technique are analyzed. Several standard datasets are
used to evaluate the performance of the developed technique. A comparative
study shows that the developed technique is superior.
| [
{
"version": "v1",
"created": "Thu, 1 Sep 2016 01:50:18 GMT"
},
{
"version": "v2",
"created": "Sun, 22 Jan 2017 09:52:06 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Venkatesan",
"Rajasekar",
""
],
[
"Er",
"Meng Joo",
""
]
] | TITLE: A Novel Progressive Learning Technique for Multi-class Classification
ABSTRACT: In this paper, a progressive learning technique for multi-class
classification is proposed. This newly developed learning technique is
independent of the number of class constraints and it can learn new classes
while still retaining the knowledge of previous classes. Whenever a new class
(non-native to the knowledge learnt thus far) is encountered, the neural
network structure gets remodeled automatically by facilitating new neurons and
interconnections, and the parameters are calculated in such a way that it
retains the knowledge learnt thus far. This technique is suitable for
real-world applications where the number of classes is often unknown and online
learning from real-time data is required. The consistency and the complexity of
the progressive learning technique are analyzed. Several standard datasets are
used to evaluate the performance of the developed technique. A comparative
study shows that the developed technique is superior.
| no_new_dataset | 0.948632 |
1611.08749 | Randall Balestriero | Herve Glotin, Julien Ricard, Randall Balestriero | Fast Chirplet Transform to Enhance CNN Machine Listening - Validation on
Animal calls and Speech | null | null | null | null | cs.SD | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The scattering framework offers an optimal hierarchical convolutional
decomposition according to its kernels. Convolutional Neural Net (CNN) can be
seen as an optimal kernel decomposition, nevertheless it requires large amount
of training data to learn its kernels. We propose a trade-off between these two
approaches: a Chirplet kernel as an efficient Q constant bioacoustic
representation to pretrain CNN. First we motivate Chirplet bioinspired auditory
representation. Second we give the first algorithm (and code) of a Fast
Chirplet Transform (FCT). Third, we demonstrate the computation efficiency of
FCT on large environmental data base: months of Orca recordings, and 1000 Birds
species from the LifeClef challenge. Fourth, we validate FCT on the vowels
subset of the Speech TIMIT dataset. The results show that FCT accelerates CNN
when it pretrains low level layers: it reduces training duration by -28\% for
birds classification, and by -26% for vowels classification. Scores are also
enhanced by FCT pretraining, with a relative gain of +7.8% of Mean Average
Precision on birds, and +2.3\% of vowel accuracy against raw audio CNN. We
conclude on perspectives on tonotopic FCT deep machine listening, and
inter-species bioacoustic transfer learning to generalise the representation of
animal communication systems.
| [
{
"version": "v1",
"created": "Sat, 26 Nov 2016 22:16:35 GMT"
},
{
"version": "v2",
"created": "Sun, 22 Jan 2017 22:28:47 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Glotin",
"Herve",
""
],
[
"Ricard",
"Julien",
""
],
[
"Balestriero",
"Randall",
""
]
] | TITLE: Fast Chirplet Transform to Enhance CNN Machine Listening - Validation on
Animal calls and Speech
ABSTRACT: The scattering framework offers an optimal hierarchical convolutional
decomposition according to its kernels. Convolutional Neural Net (CNN) can be
seen as an optimal kernel decomposition, nevertheless it requires large amount
of training data to learn its kernels. We propose a trade-off between these two
approaches: a Chirplet kernel as an efficient Q constant bioacoustic
representation to pretrain CNN. First we motivate Chirplet bioinspired auditory
representation. Second we give the first algorithm (and code) of a Fast
Chirplet Transform (FCT). Third, we demonstrate the computation efficiency of
FCT on large environmental data base: months of Orca recordings, and 1000 Birds
species from the LifeClef challenge. Fourth, we validate FCT on the vowels
subset of the Speech TIMIT dataset. The results show that FCT accelerates CNN
when it pretrains low level layers: it reduces training duration by -28\% for
birds classification, and by -26% for vowels classification. Scores are also
enhanced by FCT pretraining, with a relative gain of +7.8% of Mean Average
Precision on birds, and +2.3\% of vowel accuracy against raw audio CNN. We
conclude on perspectives on tonotopic FCT deep machine listening, and
inter-species bioacoustic transfer learning to generalise the representation of
animal communication systems.
| no_new_dataset | 0.950319 |
1612.06027 | Ryan Cotterell Ryan D Cotterell | Katharina Kann and Ryan Cotterell and Hinrich Sch\"utze | Neural Multi-Source Morphological Reinflection | Accepted at EACL 2017. Camera Ready Version | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We explore the task of multi-source morphological reinflection, which
generalizes the standard, single-source version. The input consists of (i) a
target tag and (ii) multiple pairs of source form and source tag for a lemma.
The motivation is that it is beneficial to have access to more than one source
form since different source forms can provide complementary information, e.g.,
different stems. We further present a novel extension to the encoder- decoder
recurrent neural architecture, consisting of multiple encoders, to better solve
the task. We show that our new architecture outperforms single-source
reinflection models and publish our dataset for multi-source morphological
reinflection to facilitate future research.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2016 02:21:24 GMT"
},
{
"version": "v2",
"created": "Mon, 26 Dec 2016 06:22:45 GMT"
},
{
"version": "v3",
"created": "Sun, 22 Jan 2017 09:30:10 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Kann",
"Katharina",
""
],
[
"Cotterell",
"Ryan",
""
],
[
"Schütze",
"Hinrich",
""
]
] | TITLE: Neural Multi-Source Morphological Reinflection
ABSTRACT: We explore the task of multi-source morphological reinflection, which
generalizes the standard, single-source version. The input consists of (i) a
target tag and (ii) multiple pairs of source form and source tag for a lemma.
The motivation is that it is beneficial to have access to more than one source
form since different source forms can provide complementary information, e.g.,
different stems. We further present a novel extension to the encoder- decoder
recurrent neural architecture, consisting of multiple encoders, to better solve
the task. We show that our new architecture outperforms single-source
reinflection models and publish our dataset for multi-source morphological
reinflection to facilitate future research.
| new_dataset | 0.947962 |
1701.05923 | Fathi Salem | Rahul Dey and Fathi M. Salem | Gate-Variants of Gated Recurrent Unit (GRU) Neural Networks | 5 pages, 8 Figures, 4 Tables | null | null | null | cs.NE stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The paper evaluates three variants of the Gated Recurrent Unit (GRU) in
recurrent neural networks (RNN) by reducing parameters in the update and reset
gates. We evaluate the three variant GRU models on MNIST and IMDB datasets and
show that these GRU-RNN variant models perform as well as the original GRU RNN
model while reducing the computational expense.
| [
{
"version": "v1",
"created": "Fri, 20 Jan 2017 20:53:51 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Dey",
"Rahul",
""
],
[
"Salem",
"Fathi M.",
""
]
] | TITLE: Gate-Variants of Gated Recurrent Unit (GRU) Neural Networks
ABSTRACT: The paper evaluates three variants of the Gated Recurrent Unit (GRU) in
recurrent neural networks (RNN) by reducing parameters in the update and reset
gates. We evaluate the three variant GRU models on MNIST and IMDB datasets and
show that these GRU-RNN variant models perform as well as the original GRU RNN
model while reducing the computational expense.
| no_new_dataset | 0.953751 |
1701.05982 | Sudhakar Singh | Sudhakar Singh, Rakhi Garg, P. K. Mishra | Observations on Factors Affecting Performance of MapReduce based Apriori
on Hadoop Cluster | 8 pages, 8 figures, International Conference on Computing,
Communication and Automation (ICCCA2016) | 2016 International Conference on Computing, Communication and
Automation (ICCCA), Greater Noida, India, 2016, pp. 87-94 | 10.1109/CCAA.2016.7813695 | 466 | cs.DB cs.DC cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Designing fast and scalable algorithm for mining frequent itemsets is always
being a most eminent and promising problem of data mining. Apriori is one of
the most broadly used and popular algorithm of frequent itemset mining.
Designing efficient algorithms on MapReduce framework to process and analyze
big datasets is contemporary research nowadays. In this paper, we have focused
on the performance of MapReduce based Apriori on homogeneous as well as on
heterogeneous Hadoop cluster. We have investigated a number of factors that
significantly affects the execution time of MapReduce based Apriori running on
homogeneous and heterogeneous Hadoop Cluster. Factors are specific to both
algorithmic and non-algorithmic improvements. Considered factors specific to
algorithmic improvements are filtered transactions and data structures.
Experimental results show that how an appropriate data structure and filtered
transactions technique drastically reduce the execution time. The
non-algorithmic factors include speculative execution, nodes with poor
performance, data locality & distribution of data blocks, and parallelism
control with input split size. We have applied strategies against these factors
and fine tuned the relevant parameters in our particular application.
Experimental results show that if cluster specific parameters are taken care of
then there is a significant reduction in execution time. Also we have discussed
the issues regarding MapReduce implementation of Apriori which may
significantly influence the performance.
| [
{
"version": "v1",
"created": "Sat, 21 Jan 2017 05:12:13 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Singh",
"Sudhakar",
""
],
[
"Garg",
"Rakhi",
""
],
[
"Mishra",
"P. K.",
""
]
] | TITLE: Observations on Factors Affecting Performance of MapReduce based Apriori
on Hadoop Cluster
ABSTRACT: Designing fast and scalable algorithm for mining frequent itemsets is always
being a most eminent and promising problem of data mining. Apriori is one of
the most broadly used and popular algorithm of frequent itemset mining.
Designing efficient algorithms on MapReduce framework to process and analyze
big datasets is contemporary research nowadays. In this paper, we have focused
on the performance of MapReduce based Apriori on homogeneous as well as on
heterogeneous Hadoop cluster. We have investigated a number of factors that
significantly affects the execution time of MapReduce based Apriori running on
homogeneous and heterogeneous Hadoop Cluster. Factors are specific to both
algorithmic and non-algorithmic improvements. Considered factors specific to
algorithmic improvements are filtered transactions and data structures.
Experimental results show that how an appropriate data structure and filtered
transactions technique drastically reduce the execution time. The
non-algorithmic factors include speculative execution, nodes with poor
performance, data locality & distribution of data blocks, and parallelism
control with input split size. We have applied strategies against these factors
and fine tuned the relevant parameters in our particular application.
Experimental results show that if cluster specific parameters are taken care of
then there is a significant reduction in execution time. Also we have discussed
the issues regarding MapReduce implementation of Apriori which may
significantly influence the performance.
| no_new_dataset | 0.946646 |
1701.06075 | Linhong Zhu | Dingxiong Deng, Fan Bai, Yiqi Tang, Shuigeng Zhou, Cyrus Shahabi,
Linhong Zhu | Label Propagation on K-partite Graphs with Heterophily | null | null | null | null | cs.LG cs.AI cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, for the first time, we study label propagation in
heterogeneous graphs under heterophily assumption. Homophily label propagation
(i.e., two connected nodes share similar labels) in homogeneous graph (with
same types of vertices and relations) has been extensively studied before.
Unfortunately, real-life networks are heterogeneous, they contain different
types of vertices (e.g., users, images, texts) and relations (e.g.,
friendships, co-tagging) and allow for each node to propagate both the same and
opposite copy of labels to its neighbors. We propose a $\mathcal{K}$-partite
label propagation model to handle the mystifying combination of heterogeneous
nodes/relations and heterophily propagation. With this model, we develop a
novel label inference algorithm framework with update rules in near-linear time
complexity. Since real networks change over time, we devise an incremental
approach, which supports fast updates for both new data and evidence (e.g.,
ground truth labels) with guaranteed efficiency. We further provide a utility
function to automatically determine whether an incremental or a re-modeling
approach is favored. Extensive experiments on real datasets have verified the
effectiveness and efficiency of our approach, and its superiority over the
state-of-the-art label propagation methods.
| [
{
"version": "v1",
"created": "Sat, 21 Jan 2017 19:47:38 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Deng",
"Dingxiong",
""
],
[
"Bai",
"Fan",
""
],
[
"Tang",
"Yiqi",
""
],
[
"Zhou",
"Shuigeng",
""
],
[
"Shahabi",
"Cyrus",
""
],
[
"Zhu",
"Linhong",
""
]
] | TITLE: Label Propagation on K-partite Graphs with Heterophily
ABSTRACT: In this paper, for the first time, we study label propagation in
heterogeneous graphs under heterophily assumption. Homophily label propagation
(i.e., two connected nodes share similar labels) in homogeneous graph (with
same types of vertices and relations) has been extensively studied before.
Unfortunately, real-life networks are heterogeneous, they contain different
types of vertices (e.g., users, images, texts) and relations (e.g.,
friendships, co-tagging) and allow for each node to propagate both the same and
opposite copy of labels to its neighbors. We propose a $\mathcal{K}$-partite
label propagation model to handle the mystifying combination of heterogeneous
nodes/relations and heterophily propagation. With this model, we develop a
novel label inference algorithm framework with update rules in near-linear time
complexity. Since real networks change over time, we devise an incremental
approach, which supports fast updates for both new data and evidence (e.g.,
ground truth labels) with guaranteed efficiency. We further provide a utility
function to automatically determine whether an incremental or a re-modeling
approach is favored. Extensive experiments on real datasets have verified the
effectiveness and efficiency of our approach, and its superiority over the
state-of-the-art label propagation methods.
| no_new_dataset | 0.951459 |
1701.06207 | Akash Das Sarma | Ayush Jain, Akash Das Sarma, Aditya Parameswaran, Jennifer Widom | Understanding Workers, Developing Effective Tasks, and Enhancing
Marketplace Dynamics: A Study of a Large Crowdsourcing Marketplace | null | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We conduct an experimental analysis of a dataset comprising over 27 million
microtasks performed by over 70,000 workers issued to a large crowdsourcing
marketplace between 2012-2016. Using this data---never before analyzed in an
academic context---we shed light on three crucial aspects of crowdsourcing: (1)
Task design --- helping requesters understand what constitutes an effective
task, and how to go about designing one; (2) Marketplace dynamics --- helping
marketplace administrators and designers understand the interaction between
tasks and workers, and the corresponding marketplace load; and (3) Worker
behavior --- understanding worker attention spans, lifetimes, and general
behavior, for the improvement of the crowdsourcing ecosystem as a whole.
| [
{
"version": "v1",
"created": "Sun, 22 Jan 2017 19:04:27 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Jain",
"Ayush",
""
],
[
"Sarma",
"Akash Das",
""
],
[
"Parameswaran",
"Aditya",
""
],
[
"Widom",
"Jennifer",
""
]
] | TITLE: Understanding Workers, Developing Effective Tasks, and Enhancing
Marketplace Dynamics: A Study of a Large Crowdsourcing Marketplace
ABSTRACT: We conduct an experimental analysis of a dataset comprising over 27 million
microtasks performed by over 70,000 workers issued to a large crowdsourcing
marketplace between 2012-2016. Using this data---never before analyzed in an
academic context---we shed light on three crucial aspects of crowdsourcing: (1)
Task design --- helping requesters understand what constitutes an effective
task, and how to go about designing one; (2) Marketplace dynamics --- helping
marketplace administrators and designers understand the interaction between
tasks and workers, and the corresponding marketplace load; and (3) Worker
behavior --- understanding worker attention spans, lifetimes, and general
behavior, for the improvement of the crowdsourcing ecosystem as a whole.
| no_new_dataset | 0.675765 |
1701.06225 | Omar Montasser | Omar Montasser and Daniel Kifer | Predicting Demographics of High-Resolution Geographies with Geotagged
Tweets | 6 pages, AAAI-17 preprint | null | null | null | cs.LG cs.SI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider the problem of predicting demographics of
geographic units given geotagged Tweets that are composed within these units.
Traditional survey methods that offer demographics estimates are usually
limited in terms of geographic resolution, geographic boundaries, and time
intervals. Thus, it would be highly useful to develop computational methods
that can complement traditional survey methods by offering demographics
estimates at finer geographic resolutions, with flexible geographic boundaries
(i.e. not confined to administrative boundaries), and at different time
intervals. While prior work has focused on predicting demographics and health
statistics at relatively coarse geographic resolutions such as the county-level
or state-level, we introduce an approach to predict demographics at finer
geographic resolutions such as the blockgroup-level. For the task of predicting
gender and race/ethnicity counts at the blockgroup-level, an approach adapted
from prior work to our problem achieves an average correlation of 0.389
(gender) and 0.569 (race) on a held-out test dataset. Our approach outperforms
this prior approach with an average correlation of 0.671 (gender) and 0.692
(race).
| [
{
"version": "v1",
"created": "Sun, 22 Jan 2017 22:16:46 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Montasser",
"Omar",
""
],
[
"Kifer",
"Daniel",
""
]
] | TITLE: Predicting Demographics of High-Resolution Geographies with Geotagged
Tweets
ABSTRACT: In this paper, we consider the problem of predicting demographics of
geographic units given geotagged Tweets that are composed within these units.
Traditional survey methods that offer demographics estimates are usually
limited in terms of geographic resolution, geographic boundaries, and time
intervals. Thus, it would be highly useful to develop computational methods
that can complement traditional survey methods by offering demographics
estimates at finer geographic resolutions, with flexible geographic boundaries
(i.e. not confined to administrative boundaries), and at different time
intervals. While prior work has focused on predicting demographics and health
statistics at relatively coarse geographic resolutions such as the county-level
or state-level, we introduce an approach to predict demographics at finer
geographic resolutions such as the blockgroup-level. For the task of predicting
gender and race/ethnicity counts at the blockgroup-level, an approach adapted
from prior work to our problem achieves an average correlation of 0.389
(gender) and 0.569 (race) on a held-out test dataset. Our approach outperforms
this prior approach with an average correlation of 0.671 (gender) and 0.692
(race).
| no_new_dataset | 0.946101 |
1701.06247 | Hongjie Shi | Hongjie Shi, Takashi Ushio, Mitsuru Endo, Katsuyoshi Yamagami, Noriaki
Horii | A Multichannel Convolutional Neural Network For Cross-language Dialog
State Tracking | Copyright 2016 IEEE. Published in the 2016 IEEE Workshop on Spoken
Language Technology (SLT 2016) | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The fifth Dialog State Tracking Challenge (DSTC5) introduces a new
cross-language dialog state tracking scenario, where the participants are asked
to build their trackers based on the English training corpus, while evaluating
them with the unlabeled Chinese corpus. Although the computer-generated
translations for both English and Chinese corpus are provided in the dataset,
these translations contain errors and careless use of them can easily hurt the
performance of the built trackers. To address this problem, we propose a
multichannel Convolutional Neural Networks (CNN) architecture, in which we
treat English and Chinese language as different input channels of one single
CNN model. In the evaluation of DSTC5, we found that such multichannel
architecture can effectively improve the robustness against translation errors.
Additionally, our method for DSTC5 is purely machine learning based and
requires no prior knowledge about the target language. We consider this a
desirable property for building a tracker in the cross-language context, as not
every developer will be familiar with both languages.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 01:36:10 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Shi",
"Hongjie",
""
],
[
"Ushio",
"Takashi",
""
],
[
"Endo",
"Mitsuru",
""
],
[
"Yamagami",
"Katsuyoshi",
""
],
[
"Horii",
"Noriaki",
""
]
] | TITLE: A Multichannel Convolutional Neural Network For Cross-language Dialog
State Tracking
ABSTRACT: The fifth Dialog State Tracking Challenge (DSTC5) introduces a new
cross-language dialog state tracking scenario, where the participants are asked
to build their trackers based on the English training corpus, while evaluating
them with the unlabeled Chinese corpus. Although the computer-generated
translations for both English and Chinese corpus are provided in the dataset,
these translations contain errors and careless use of them can easily hurt the
performance of the built trackers. To address this problem, we propose a
multichannel Convolutional Neural Networks (CNN) architecture, in which we
treat English and Chinese language as different input channels of one single
CNN model. In the evaluation of DSTC5, we found that such multichannel
architecture can effectively improve the robustness against translation errors.
Additionally, our method for DSTC5 is purely machine learning based and
requires no prior knowledge about the target language. We consider this a
desirable property for building a tracker in the cross-language context, as not
every developer will be familiar with both languages.
| no_new_dataset | 0.943815 |
1701.06276 | Georgios Stylianou | Georgios Stylianou | Stay-point Identification as Curve Extrema | null | null | null | null | cs.OH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In a nutshell, stay-points are locations that a person has stopped for some
amount of time. Previous work depends mainly on stay-point identification
methods using experimentally fine tuned threshold values. These behave well on
their experimental datasets but may exhibit reduced performance on other
datasets.
In this work, we demonstrate the potential of a geometry-based method for
stay-point extraction. This is accomplished by transforming the user's
trajectory path to a two-dimensional discrete time series curve that in turn
transforms the stay-points to the local minima of the first derivative of this
curve.
To demonstrate the soundness of the proposed method, we evaluated it on raw,
noisy trajectory data acquired over the period of 28 different days using four
different techniques. The results demonstrate, among others, that given a good
trajectory tracking technique, we can identify correctly 86% to 98% of the
stay-points.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 06:45:01 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Stylianou",
"Georgios",
""
]
] | TITLE: Stay-point Identification as Curve Extrema
ABSTRACT: In a nutshell, stay-points are locations that a person has stopped for some
amount of time. Previous work depends mainly on stay-point identification
methods using experimentally fine tuned threshold values. These behave well on
their experimental datasets but may exhibit reduced performance on other
datasets.
In this work, we demonstrate the potential of a geometry-based method for
stay-point extraction. This is accomplished by transforming the user's
trajectory path to a two-dimensional discrete time series curve that in turn
transforms the stay-points to the local minima of the first derivative of this
curve.
To demonstrate the soundness of the proposed method, we evaluated it on raw,
noisy trajectory data acquired over the period of 28 different days using four
different techniques. The results demonstrate, among others, that given a good
trajectory tracking technique, we can identify correctly 86% to 98% of the
stay-points.
| no_new_dataset | 0.944995 |
1701.06439 | Teik Koon Cheang | Teik Koon Cheang, Yong Shean Chong, Yong Haur Tay | Segmentation-free Vehicle License Plate Recognition using ConvNet-RNN | 5 pages, 3 figures, International Workshop on Advanced Image
Technology, January, 8-10, 2017. Penang, Malaysia. Proceeding IWAIT2017 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While vehicle license plate recognition (VLPR) is usually done with a sliding
window approach, it can have limited performance on datasets with characters
that are of variable width. This can be solved by hand-crafting algorithms to
prescale the characters. While this approach can work fairly well, the
recognizer is only aware of the pixels within each detector window, and fails
to account for other contextual information that might be present in other
parts of the image. A sliding window approach also requires training data in
the form of presegmented characters, which can be more difficult to obtain. In
this paper, we propose a unified ConvNet-RNN model to recognize real-world
captured license plate photographs. By using a Convolutional Neural Network
(ConvNet) to perform feature extraction and using a Recurrent Neural Network
(RNN) for sequencing, we address the problem of sliding window approaches being
unable to access the context of the entire image by feeding the entire image as
input to the ConvNet. This has the added benefit of being able to perform
end-to-end training of the entire model on labelled, full license plate images.
Experimental results comparing the ConvNet-RNN architecture to a sliding
window-based approach shows that the ConvNet-RNN architecture performs
significantly better.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 15:11:12 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Cheang",
"Teik Koon",
""
],
[
"Chong",
"Yong Shean",
""
],
[
"Tay",
"Yong Haur",
""
]
] | TITLE: Segmentation-free Vehicle License Plate Recognition using ConvNet-RNN
ABSTRACT: While vehicle license plate recognition (VLPR) is usually done with a sliding
window approach, it can have limited performance on datasets with characters
that are of variable width. This can be solved by hand-crafting algorithms to
prescale the characters. While this approach can work fairly well, the
recognizer is only aware of the pixels within each detector window, and fails
to account for other contextual information that might be present in other
parts of the image. A sliding window approach also requires training data in
the form of presegmented characters, which can be more difficult to obtain. In
this paper, we propose a unified ConvNet-RNN model to recognize real-world
captured license plate photographs. By using a Convolutional Neural Network
(ConvNet) to perform feature extraction and using a Recurrent Neural Network
(RNN) for sequencing, we address the problem of sliding window approaches being
unable to access the context of the entire image by feeding the entire image as
input to the ConvNet. This has the added benefit of being able to perform
end-to-end training of the entire model on labelled, full license plate images.
Experimental results comparing the ConvNet-RNN architecture to a sliding
window-based approach shows that the ConvNet-RNN architecture performs
significantly better.
| no_new_dataset | 0.948202 |
1701.06450 | Andrea Baisero | Andrea Baisero, Stefan Otte, Peter Englert and Marc Toussaint | Identification of Unmodeled Objects from Symbolic Descriptions | null | null | null | null | stat.ML cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Successful human-robot cooperation hinges on each agent's ability to process
and exchange information about the shared environment and the task at hand.
Human communication is primarily based on symbolic abstractions of object
properties, rather than precise quantitative measures. A comprehensive robotic
framework thus requires an integrated communication module which is able to
establish a link and convert between perceptual and abstract information.
The ability to interpret composite symbolic descriptions enables an
autonomous agent to a) operate in unstructured and cluttered environments, in
tasks which involve unmodeled or never seen before objects; and b) exploit the
aggregation of multiple symbolic properties as an instance of ensemble
learning, to improve identification performance even when the individual
predicates encode generic information or are imprecisely grounded.
We propose a discriminative probabilistic model which interprets symbolic
descriptions to identify the referent object contextually w.r.t.\ the structure
of the environment and other objects. The model is trained using a collected
dataset of identifications, and its performance is evaluated by quantitative
measures and a live demo developed on the PR2 robot platform, which integrates
elements of perception, object extraction, object identification and grasping.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 15:26:01 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Baisero",
"Andrea",
""
],
[
"Otte",
"Stefan",
""
],
[
"Englert",
"Peter",
""
],
[
"Toussaint",
"Marc",
""
]
] | TITLE: Identification of Unmodeled Objects from Symbolic Descriptions
ABSTRACT: Successful human-robot cooperation hinges on each agent's ability to process
and exchange information about the shared environment and the task at hand.
Human communication is primarily based on symbolic abstractions of object
properties, rather than precise quantitative measures. A comprehensive robotic
framework thus requires an integrated communication module which is able to
establish a link and convert between perceptual and abstract information.
The ability to interpret composite symbolic descriptions enables an
autonomous agent to a) operate in unstructured and cluttered environments, in
tasks which involve unmodeled or never seen before objects; and b) exploit the
aggregation of multiple symbolic properties as an instance of ensemble
learning, to improve identification performance even when the individual
predicates encode generic information or are imprecisely grounded.
We propose a discriminative probabilistic model which interprets symbolic
descriptions to identify the referent object contextually w.r.t.\ the structure
of the environment and other objects. The model is trained using a collected
dataset of identifications, and its performance is evaluated by quantitative
measures and a live demo developed on the PR2 robot platform, which integrates
elements of perception, object extraction, object identification and grasping.
| no_new_dataset | 0.943348 |
1701.06462 | Eu Koon Cheang | Eu Koon Cheang, Teik Koon Cheang, Yong Haur Tay | Using Convolutional Neural Networks to Count Palm Trees in Satellite
Images | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we propose a supervised learning system for counting and
localizing palm trees in high-resolution, panchromatic satellite imagery
(40cm/pixel to 1.5m/pixel). A convolutional neural network classifier trained
on a set of palm and no-palm images is applied across a satellite image scene
in a sliding window fashion. The resultant confidence map is smoothed with a
uniform filter. A non-maximal suppression is applied onto the smoothed
confidence map to obtain peaks. Trained with a small dataset of 500 images of
size 40x40 cropped from satellite images, the system manages to achieve a tree
count accuracy of over 99%.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2017 15:38:52 GMT"
}
] | 2017-01-24T00:00:00 | [
[
"Cheang",
"Eu Koon",
""
],
[
"Cheang",
"Teik Koon",
""
],
[
"Tay",
"Yong Haur",
""
]
] | TITLE: Using Convolutional Neural Networks to Count Palm Trees in Satellite
Images
ABSTRACT: In this paper we propose a supervised learning system for counting and
localizing palm trees in high-resolution, panchromatic satellite imagery
(40cm/pixel to 1.5m/pixel). A convolutional neural network classifier trained
on a set of palm and no-palm images is applied across a satellite image scene
in a sliding window fashion. The resultant confidence map is smoothed with a
uniform filter. A non-maximal suppression is applied onto the smoothed
confidence map to obtain peaks. Trained with a small dataset of 500 images of
size 40x40 cropped from satellite images, the system manages to achieve a tree
count accuracy of over 99%.
| no_new_dataset | 0.627352 |
1609.04214 | Shujun Li Dr. | Aamo Iorliam, Santosh Tirunagari, Anthony T.S. Ho, Shujun Li, Adrian
Waller and Norman Poh | "Flow Size Difference" Can Make a Difference: Detecting Malicious TCP
Network Flows Based on Benford's Law | 13 pages, 3 figures | null | null | null | cs.CR cs.AI cs.NI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Statistical characteristics of network traffic have attracted a significant
amount of research for automated network intrusion detection, some of which
looked at applications of natural statistical laws such as Zipf's law,
Benford's law and the Pareto distribution. In this paper, we present the
application of Benford's law to a new network flow metric "flow size
difference", which have not been studied before by other researchers, to build
an unsupervised flow-based intrusion detection system (IDS). The method was
inspired by our observation on a large number of TCP flow datasets where normal
flows tend to follow Benford's law closely but malicious flows tend to deviate
significantly from it. The proposed IDS is unsupervised, so it can be easily
deployed without any training. It has two simple operational parameters with a
clear semantic meaning, allowing the IDS operator to set and adapt their values
intuitively to adjust the overall performance of the IDS. We tested the
proposed IDS on two (one closed and one public) datasets, and proved its
efficiency in terms of AUC (area under the ROC curve). Our work showed the
"flow size difference" has a great potential to improve the performance of any
flow-based network IDSs.
| [
{
"version": "v1",
"created": "Wed, 14 Sep 2016 10:51:00 GMT"
},
{
"version": "v2",
"created": "Fri, 20 Jan 2017 18:22:47 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Iorliam",
"Aamo",
""
],
[
"Tirunagari",
"Santosh",
""
],
[
"Ho",
"Anthony T. S.",
""
],
[
"Li",
"Shujun",
""
],
[
"Waller",
"Adrian",
""
],
[
"Poh",
"Norman",
""
]
] | TITLE: "Flow Size Difference" Can Make a Difference: Detecting Malicious TCP
Network Flows Based on Benford's Law
ABSTRACT: Statistical characteristics of network traffic have attracted a significant
amount of research for automated network intrusion detection, some of which
looked at applications of natural statistical laws such as Zipf's law,
Benford's law and the Pareto distribution. In this paper, we present the
application of Benford's law to a new network flow metric "flow size
difference", which have not been studied before by other researchers, to build
an unsupervised flow-based intrusion detection system (IDS). The method was
inspired by our observation on a large number of TCP flow datasets where normal
flows tend to follow Benford's law closely but malicious flows tend to deviate
significantly from it. The proposed IDS is unsupervised, so it can be easily
deployed without any training. It has two simple operational parameters with a
clear semantic meaning, allowing the IDS operator to set and adapt their values
intuitively to adjust the overall performance of the IDS. We tested the
proposed IDS on two (one closed and one public) datasets, and proved its
efficiency in terms of AUC (area under the ROC curve). Our work showed the
"flow size difference" has a great potential to improve the performance of any
flow-based network IDSs.
| no_new_dataset | 0.950365 |
1701.05581 | Diptesh Kanojia | Abhijit Mishra, Diptesh Kanojia, Seema Nagar, Kuntal Dey and Pushpak
Bhattacharyya | Leveraging Cognitive Features for Sentiment Analysis | The SIGNLL Conference on Computational Natural Language Learning
(CoNLL 2016) | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sentiments expressed in user-generated short text and sentences are nuanced
by subtleties at lexical, syntactic, semantic and pragmatic levels. To address
this, we propose to augment traditional features used for sentiment analysis
and sarcasm detection, with cognitive features derived from the eye-movement
patterns of readers. Statistical classification using our enhanced feature set
improves the performance (F-score) of polarity detection by a maximum of 3.7%
and 9.3% on two datasets, over the systems that use only traditional features.
We perform feature significance analysis, and experiment on a held-out dataset,
showing that cognitive features indeed empower sentiment analyzers to handle
complex constructs.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 19:58:26 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Mishra",
"Abhijit",
""
],
[
"Kanojia",
"Diptesh",
""
],
[
"Nagar",
"Seema",
""
],
[
"Dey",
"Kuntal",
""
],
[
"Bhattacharyya",
"Pushpak",
""
]
] | TITLE: Leveraging Cognitive Features for Sentiment Analysis
ABSTRACT: Sentiments expressed in user-generated short text and sentences are nuanced
by subtleties at lexical, syntactic, semantic and pragmatic levels. To address
this, we propose to augment traditional features used for sentiment analysis
and sarcasm detection, with cognitive features derived from the eye-movement
patterns of readers. Statistical classification using our enhanced feature set
improves the performance (F-score) of polarity detection by a maximum of 3.7%
and 9.3% on two datasets, over the systems that use only traditional features.
We perform feature significance analysis, and experiment on a held-out dataset,
showing that cognitive features indeed empower sentiment analyzers to handle
complex constructs.
| no_new_dataset | 0.935759 |
1701.05595 | Mohammad Mahmoodi | Mohammad Reza Mahmoodi | Fast and Efficient Skin Detection for Facial Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, an efficient skin detection system is proposed. The algorithm
is based on a very fast efficient pre-processing step utilizing the concept of
ternary conversion in order to identify candidate windows and subsequently, a
novel local two-stage diffusion method which has F-score accuracy of 0.5978 on
SDD dataset. The pre-processing step has been proven to be useful to boost the
speed of the system by eliminating 82% of an image in average. This is obtained
by keeping the true positive rate above 98%. In addition, a novel segmentation
algorithm is also designed to process candidate windows which is quantitatively
and qualitatively proven to be very efficient in term of accuracy. The
algorithm has been implemented in FPGA to obtain real-time processing speed.
The system is designed fully pipeline and the inherent parallel structure of
the algorithm is fully exploited to maximize the performance. The system is
implemented on a Spartan-6 LXT45 Xilinx FPGA and it is capable of processing 98
frames of 640*480 24-bit color images per second.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 20:43:27 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Mahmoodi",
"Mohammad Reza",
""
]
] | TITLE: Fast and Efficient Skin Detection for Facial Detection
ABSTRACT: In this paper, an efficient skin detection system is proposed. The algorithm
is based on a very fast efficient pre-processing step utilizing the concept of
ternary conversion in order to identify candidate windows and subsequently, a
novel local two-stage diffusion method which has F-score accuracy of 0.5978 on
SDD dataset. The pre-processing step has been proven to be useful to boost the
speed of the system by eliminating 82% of an image in average. This is obtained
by keeping the true positive rate above 98%. In addition, a novel segmentation
algorithm is also designed to process candidate windows which is quantitatively
and qualitatively proven to be very efficient in term of accuracy. The
algorithm has been implemented in FPGA to obtain real-time processing speed.
The system is designed fully pipeline and the inherent parallel structure of
the algorithm is fully exploited to maximize the performance. The system is
implemented on a Spartan-6 LXT45 Xilinx FPGA and it is capable of processing 98
frames of 640*480 24-bit color images per second.
| no_new_dataset | 0.949295 |
1701.05596 | Roger Schaer | Dimitrios Markonis, Roger Schaer, Alba Garc\'ia Seco de Herrera,
Henning M\"uller | The Parallel Distributed Image Search Engine (ParaDISE) | 23 pages, 9 figures | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image retrieval is a complex task that differs according to the context and
the user requirements in any specific field, for example in a medical
environment. Search by text is often not possible or optimal and retrieval by
the visual content does not always succeed in modelling high-level concepts
that a user is looking for. Modern image retrieval techniques consist of
multiple steps and aim to retrieve information from large--scale datasets and
not only based on global image appearance but local features and if possible in
a connection between visual features and text or semantics. This paper presents
the Parallel Distributed Image Search Engine (ParaDISE), an image retrieval
system that combines visual search with text--based retrieval and that is
available as open source and free of charge. The main design concepts of
ParaDISE are flexibility, expandability, scalability and interoperability.
These concepts constitute the system, able to be used both in real-world
applications and as an image retrieval research platform. Apart from the
architecture and the implementation of the system, two use cases are described,
an application of ParaDISE in retrieval of images from the medical literature
and a visual feature evaluation for medical image retrieval. Future steps
include the creation of an open source community that will contribute and
expand this platform based on the existing parts.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 20:51:56 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Markonis",
"Dimitrios",
""
],
[
"Schaer",
"Roger",
""
],
[
"de Herrera",
"Alba García Seco",
""
],
[
"Müller",
"Henning",
""
]
] | TITLE: The Parallel Distributed Image Search Engine (ParaDISE)
ABSTRACT: Image retrieval is a complex task that differs according to the context and
the user requirements in any specific field, for example in a medical
environment. Search by text is often not possible or optimal and retrieval by
the visual content does not always succeed in modelling high-level concepts
that a user is looking for. Modern image retrieval techniques consist of
multiple steps and aim to retrieve information from large--scale datasets and
not only based on global image appearance but local features and if possible in
a connection between visual features and text or semantics. This paper presents
the Parallel Distributed Image Search Engine (ParaDISE), an image retrieval
system that combines visual search with text--based retrieval and that is
available as open source and free of charge. The main design concepts of
ParaDISE are flexibility, expandability, scalability and interoperability.
These concepts constitute the system, able to be used both in real-world
applications and as an image retrieval research platform. Apart from the
architecture and the implementation of the system, two use cases are described,
an application of ParaDISE in retrieval of images from the medical literature
and a visual feature evaluation for medical image retrieval. Future steps
include the creation of an open source community that will contribute and
expand this platform based on the existing parts.
| no_new_dataset | 0.949342 |
1701.05632 | Simon Angus | Klaus Ackermann, Simon D Angus, Paul A Raschky | The Internet as Quantitative Social Science Platform: Insights from a
Trillion Observations | 40 pages, including 4 main figures, and appendix | null | null | null | q-fin.EC cs.CY cs.SI physics.soc-ph stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the large-scale penetration of the internet, for the first time,
humanity has become linked by a single, open, communications platform.
Harnessing this fact, we report insights arising from a unified internet
activity and location dataset of an unparalleled scope and accuracy drawn from
over a trillion (1.5$\times 10^{12}$) observations of end-user internet
connections, with temporal resolution of just 15min over 2006-2012. We first
apply this dataset to the expansion of the internet itself over 1,647 urban
agglomerations globally. We find that unique IP per capita counts reach
saturation at approximately one IP per three people, and take, on average, 16.1
years to achieve; eclipsing the estimated 100- and 60- year saturation times
for steam-power and electrification respectively. Next, we use intra-diurnal
internet activity features to up-scale traditional over-night sleep
observations, producing the first global estimate of over-night sleep duration
in 645 cities over 7 years. We find statistically significant variation between
continental, national and regional sleep durations including some evidence of
global sleep duration convergence. Finally, we estimate the relationship
between internet concentration and economic outcomes in 411 OECD regions and
find that the internet's expansion is associated with negative or positive
productivity gains, depending strongly on sectoral considerations. To our
knowledge, our study is the first of its kind to use online/offline activity of
the entire internet to infer social science insights, demonstrating the
unparalleled potential of the internet as a social data-science platform.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 22:35:46 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Ackermann",
"Klaus",
""
],
[
"Angus",
"Simon D",
""
],
[
"Raschky",
"Paul A",
""
]
] | TITLE: The Internet as Quantitative Social Science Platform: Insights from a
Trillion Observations
ABSTRACT: With the large-scale penetration of the internet, for the first time,
humanity has become linked by a single, open, communications platform.
Harnessing this fact, we report insights arising from a unified internet
activity and location dataset of an unparalleled scope and accuracy drawn from
over a trillion (1.5$\times 10^{12}$) observations of end-user internet
connections, with temporal resolution of just 15min over 2006-2012. We first
apply this dataset to the expansion of the internet itself over 1,647 urban
agglomerations globally. We find that unique IP per capita counts reach
saturation at approximately one IP per three people, and take, on average, 16.1
years to achieve; eclipsing the estimated 100- and 60- year saturation times
for steam-power and electrification respectively. Next, we use intra-diurnal
internet activity features to up-scale traditional over-night sleep
observations, producing the first global estimate of over-night sleep duration
in 645 cities over 7 years. We find statistically significant variation between
continental, national and regional sleep durations including some evidence of
global sleep duration convergence. Finally, we estimate the relationship
between internet concentration and economic outcomes in 411 OECD regions and
find that the internet's expansion is associated with negative or positive
productivity gains, depending strongly on sectoral considerations. To our
knowledge, our study is the first of its kind to use online/offline activity of
the entire internet to infer social science insights, demonstrating the
unparalleled potential of the internet as a social data-science platform.
| no_new_dataset | 0.932207 |
1701.05779 | Sungho Jeon | Sungho Jeon, Jong-Woo Shin, Young-Jun Lee, Woong-Hee Kim, YoungHyoun
Kwon, and Hae-Yong Yang | Empirical Study of Drone Sound Detection in Real-Life Environment with
Deep Neural Networks | IEEE 5 Pages, Submitted | null | null | null | cs.SD cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work aims to investigate the use of deep neural network to detect
commercial hobby drones in real-life environments by analyzing their sound
data. The purpose of work is to contribute to a system for detecting drones
used for malicious purposes, such as for terrorism. Specifically, we present a
method capable of detecting the presence of commercial hobby drones as a binary
classification problem based on sound event detection. We recorded the sound
produced by a few popular commercial hobby drones, and then augmented this data
with diverse environmental sound data to remedy the scarcity of drone sound
data in diverse environments. We investigated the effectiveness of
state-of-the-art event sound classification methods, i.e., a Gaussian Mixture
Model (GMM), Convolutional Neural Network (CNN), and Recurrent Neural Network
(RNN), for drone sound detection. Our empirical results, which were obtained
with a testing dataset collected on an urban street, confirmed the
effectiveness of these models for operating in a real environment. In summary,
our RNN models showed the best detection performance with an F-Score of 0.8009
with 240 ms of input audio with a short processing time, indicating their
applicability to real-time detection systems.
| [
{
"version": "v1",
"created": "Fri, 20 Jan 2017 12:48:02 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Jeon",
"Sungho",
""
],
[
"Shin",
"Jong-Woo",
""
],
[
"Lee",
"Young-Jun",
""
],
[
"Kim",
"Woong-Hee",
""
],
[
"Kwon",
"YoungHyoun",
""
],
[
"Yang",
"Hae-Yong",
""
]
] | TITLE: Empirical Study of Drone Sound Detection in Real-Life Environment with
Deep Neural Networks
ABSTRACT: This work aims to investigate the use of deep neural network to detect
commercial hobby drones in real-life environments by analyzing their sound
data. The purpose of work is to contribute to a system for detecting drones
used for malicious purposes, such as for terrorism. Specifically, we present a
method capable of detecting the presence of commercial hobby drones as a binary
classification problem based on sound event detection. We recorded the sound
produced by a few popular commercial hobby drones, and then augmented this data
with diverse environmental sound data to remedy the scarcity of drone sound
data in diverse environments. We investigated the effectiveness of
state-of-the-art event sound classification methods, i.e., a Gaussian Mixture
Model (GMM), Convolutional Neural Network (CNN), and Recurrent Neural Network
(RNN), for drone sound detection. Our empirical results, which were obtained
with a testing dataset collected on an urban street, confirmed the
effectiveness of these models for operating in a real environment. In summary,
our RNN models showed the best detection performance with an F-Score of 0.8009
with 240 ms of input audio with a short processing time, indicating their
applicability to real-time detection systems.
| no_new_dataset | 0.941868 |
1701.05818 | Nicolas Audebert | Nicolas Audebert (Palaiseau, OBELIX), Bertrand Le Saux (Palaiseau),
S\'ebastien Lef\`evre (OBELIX) | Fusion of Heterogeneous Data in Convolutional Networks for Urban
Semantic Labeling (Invited Paper) | Joint Urban Remote Sensing Event (JURSE), Mar 2017, Dubai, United
Arab Emirates. Joint Urban Remote Sensing Event 2017 | null | null | null | cs.NE cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present a novel module to perform fusion of heterogeneous
data using fully convolutional networks for semantic labeling. We introduce
residual correction as a way to learn how to fuse predictions coming out of a
dual stream architecture. Especially, we perform fusion of DSM and IRRG optical
data on the ISPRS Vaihingen dataset over a urban area and obtain new
state-of-the-art results.
| [
{
"version": "v1",
"created": "Fri, 20 Jan 2017 15:10:09 GMT"
}
] | 2017-01-23T00:00:00 | [
[
"Audebert",
"Nicolas",
"",
"Palaiseau, OBELIX"
],
[
"Saux",
"Bertrand Le",
"",
"Palaiseau"
],
[
"Lefèvre",
"Sébastien",
"",
"OBELIX"
]
] | TITLE: Fusion of Heterogeneous Data in Convolutional Networks for Urban
Semantic Labeling (Invited Paper)
ABSTRACT: In this work, we present a novel module to perform fusion of heterogeneous
data using fully convolutional networks for semantic labeling. We introduce
residual correction as a way to learn how to fuse predictions coming out of a
dual stream architecture. Especially, we perform fusion of DSM and IRRG optical
data on the ISPRS Vaihingen dataset over a urban area and obtain new
state-of-the-art results.
| no_new_dataset | 0.951278 |
1606.00061 | Jiasen Lu | Jiasen Lu, Jianwei Yang, Dhruv Batra, Devi Parikh | Hierarchical Question-Image Co-Attention for Visual Question Answering | 11 pages, 7 figures, 3 tables in 2016 Conference on Neural
Information Processing Systems (NIPS) | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A number of recent works have proposed attention models for Visual Question
Answering (VQA) that generate spatial maps highlighting image regions relevant
to answering the question. In this paper, we argue that in addition to modeling
"where to look" or visual attention, it is equally important to model "what
words to listen to" or question attention. We present a novel co-attention
model for VQA that jointly reasons about image and question attention. In
addition, our model reasons about the question (and consequently the image via
the co-attention mechanism) in a hierarchical fashion via a novel 1-dimensional
convolution neural networks (CNN). Our model improves the state-of-the-art on
the VQA dataset from 60.3% to 60.5%, and from 61.6% to 63.3% on the COCO-QA
dataset. By using ResNet, the performance is further improved to 62.1% for VQA
and 65.4% for COCO-QA.
| [
{
"version": "v1",
"created": "Tue, 31 May 2016 22:02:01 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Jun 2016 01:51:13 GMT"
},
{
"version": "v3",
"created": "Wed, 26 Oct 2016 02:15:57 GMT"
},
{
"version": "v4",
"created": "Fri, 13 Jan 2017 16:18:03 GMT"
},
{
"version": "v5",
"created": "Thu, 19 Jan 2017 05:03:33 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Lu",
"Jiasen",
""
],
[
"Yang",
"Jianwei",
""
],
[
"Batra",
"Dhruv",
""
],
[
"Parikh",
"Devi",
""
]
] | TITLE: Hierarchical Question-Image Co-Attention for Visual Question Answering
ABSTRACT: A number of recent works have proposed attention models for Visual Question
Answering (VQA) that generate spatial maps highlighting image regions relevant
to answering the question. In this paper, we argue that in addition to modeling
"where to look" or visual attention, it is equally important to model "what
words to listen to" or question attention. We present a novel co-attention
model for VQA that jointly reasons about image and question attention. In
addition, our model reasons about the question (and consequently the image via
the co-attention mechanism) in a hierarchical fashion via a novel 1-dimensional
convolution neural networks (CNN). Our model improves the state-of-the-art on
the VQA dataset from 60.3% to 60.5%, and from 61.6% to 63.3% on the COCO-QA
dataset. By using ResNet, the performance is further improved to 62.1% for VQA
and 65.4% for COCO-QA.
| no_new_dataset | 0.954774 |
1606.09075 | John V Monaco | John V. Monaco | Robust Keystroke Biometric Anomaly Detection | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Keystroke Biometrics Ongoing Competition (KBOC) presented an anomaly
detection challenge with a public keystroke dataset containing a large number
of subjects and real-world aspects. Over 300 subjects typed case-insensitive
repetitions of their first and last name, and as a result, keystroke sequences
could vary in length and order depending on the usage of modifier keys. To deal
with this, a keystroke alignment preprocessing algorithm was developed to
establish a semantic correspondence between keystrokes in mismatched sequences.
The method is robust in the sense that query keystroke sequences need only
approximately match a target sequence, and alignment is agnostic to the
particular anomaly detector used. This paper describes the fifteen
best-performing anomaly detection systems submitted to the KBOC, which ranged
from auto-encoding neural networks to ensemble methods. Manhattan distance
achieved the lowest equal error rate of 5.32%, while all fifteen systems
performed better than any other submission. Performance gains are shown to be
due in large part not to the particular anomaly detector, but to preprocessing
and score normalization techniques.
| [
{
"version": "v1",
"created": "Wed, 29 Jun 2016 13:09:29 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Jan 2017 19:19:00 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Monaco",
"John V.",
""
]
] | TITLE: Robust Keystroke Biometric Anomaly Detection
ABSTRACT: The Keystroke Biometrics Ongoing Competition (KBOC) presented an anomaly
detection challenge with a public keystroke dataset containing a large number
of subjects and real-world aspects. Over 300 subjects typed case-insensitive
repetitions of their first and last name, and as a result, keystroke sequences
could vary in length and order depending on the usage of modifier keys. To deal
with this, a keystroke alignment preprocessing algorithm was developed to
establish a semantic correspondence between keystrokes in mismatched sequences.
The method is robust in the sense that query keystroke sequences need only
approximately match a target sequence, and alignment is agnostic to the
particular anomaly detector used. This paper describes the fifteen
best-performing anomaly detection systems submitted to the KBOC, which ranged
from auto-encoding neural networks to ensemble methods. Manhattan distance
achieved the lowest equal error rate of 5.32%, while all fifteen systems
performed better than any other submission. Performance gains are shown to be
due in large part not to the particular anomaly detector, but to preprocessing
and score normalization techniques.
| no_new_dataset | 0.783906 |
1701.05105 | Zetao Chen | Zetao Chen, Adam Jacobson, Niko Sunderhauf, Ben Upcroft, Lingqiao Liu,
Chunhua Shen, Ian Reid and Michael Milford | Deep Learning Features at Scale for Visual Place Recognition | 8 pages, 10 figures. Accepted by International Conference on Robotics
and Automation (ICRA) 2017. This is the submitted version. The final
published version may be slightly different | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The success of deep learning techniques in the computer vision domain has
triggered a range of initial investigations into their utility for visual place
recognition, all using generic features from networks that were trained for
other types of recognition tasks. In this paper, we train, at large scale, two
CNN architectures for the specific place recognition task and employ a
multi-scale feature encoding method to generate condition- and
viewpoint-invariant features. To enable this training to occur, we have
developed a massive Specific PlacEs Dataset (SPED) with hundreds of examples of
place appearance change at thousands of different places, as opposed to the
semantic place type datasets currently available. This new dataset enables us
to set up a training regime that interprets place recognition as a
classification problem. We comprehensively evaluate our trained networks on
several challenging benchmark place recognition datasets and demonstrate that
they achieve an average 10% increase in performance over other place
recognition algorithms and pre-trained CNNs. By analyzing the network responses
and their differences from pre-trained networks, we provide insights into what
a network learns when training for place recognition, and what these results
signify for future research in this area.
| [
{
"version": "v1",
"created": "Wed, 18 Jan 2017 16:28:03 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Chen",
"Zetao",
""
],
[
"Jacobson",
"Adam",
""
],
[
"Sunderhauf",
"Niko",
""
],
[
"Upcroft",
"Ben",
""
],
[
"Liu",
"Lingqiao",
""
],
[
"Shen",
"Chunhua",
""
],
[
"Reid",
"Ian",
""
],
[
"Milford",
"Michael",
""
]
] | TITLE: Deep Learning Features at Scale for Visual Place Recognition
ABSTRACT: The success of deep learning techniques in the computer vision domain has
triggered a range of initial investigations into their utility for visual place
recognition, all using generic features from networks that were trained for
other types of recognition tasks. In this paper, we train, at large scale, two
CNN architectures for the specific place recognition task and employ a
multi-scale feature encoding method to generate condition- and
viewpoint-invariant features. To enable this training to occur, we have
developed a massive Specific PlacEs Dataset (SPED) with hundreds of examples of
place appearance change at thousands of different places, as opposed to the
semantic place type datasets currently available. This new dataset enables us
to set up a training regime that interprets place recognition as a
classification problem. We comprehensively evaluate our trained networks on
several challenging benchmark place recognition datasets and demonstrate that
they achieve an average 10% increase in performance over other place
recognition algorithms and pre-trained CNNs. By analyzing the network responses
and their differences from pre-trained networks, we provide insights into what
a network learns when training for place recognition, and what these results
signify for future research in this area.
| new_dataset | 0.956268 |
1701.05360 | James Booth | James Booth, Epameinondas Antonakos, Stylianos Ploumpis, George
Trigeorgis, Yannis Panagakis, and Stefanos Zafeiriou | 3D Face Morphable Models "In-the-Wild" | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D Morphable Models (3DMMs) are powerful statistical models of 3D facial
shape and texture, and among the state-of-the-art methods for reconstructing
facial shape from single images. With the advent of new 3D sensors, many 3D
facial datasets have been collected containing both neutral as well as
expressive faces. However, all datasets are captured under controlled
conditions. Thus, even though powerful 3D facial shape models can be learnt
from such data, it is difficult to build statistical texture models that are
sufficient to reconstruct faces captured in unconstrained conditions
("in-the-wild"). In this paper, we propose the first, to the best of our
knowledge, "in-the-wild" 3DMM by combining a powerful statistical model of
facial shape, which describes both identity and expression, with an
"in-the-wild" texture model. We show that the employment of such an
"in-the-wild" texture model greatly simplifies the fitting procedure, because
there is no need to optimize with regards to the illumination parameters.
Furthermore, we propose a new fast algorithm for fitting the 3DMM in arbitrary
images. Finally, we have captured the first 3D facial database with relatively
unconstrained conditions and report quantitative evaluations with
state-of-the-art performance. Complementary qualitative reconstruction results
are demonstrated on standard "in-the-wild" facial databases. An open source
implementation of our technique is released as part of the Menpo Project.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 10:27:38 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Booth",
"James",
""
],
[
"Antonakos",
"Epameinondas",
""
],
[
"Ploumpis",
"Stylianos",
""
],
[
"Trigeorgis",
"George",
""
],
[
"Panagakis",
"Yannis",
""
],
[
"Zafeiriou",
"Stefanos",
""
]
] | TITLE: 3D Face Morphable Models "In-the-Wild"
ABSTRACT: 3D Morphable Models (3DMMs) are powerful statistical models of 3D facial
shape and texture, and among the state-of-the-art methods for reconstructing
facial shape from single images. With the advent of new 3D sensors, many 3D
facial datasets have been collected containing both neutral as well as
expressive faces. However, all datasets are captured under controlled
conditions. Thus, even though powerful 3D facial shape models can be learnt
from such data, it is difficult to build statistical texture models that are
sufficient to reconstruct faces captured in unconstrained conditions
("in-the-wild"). In this paper, we propose the first, to the best of our
knowledge, "in-the-wild" 3DMM by combining a powerful statistical model of
facial shape, which describes both identity and expression, with an
"in-the-wild" texture model. We show that the employment of such an
"in-the-wild" texture model greatly simplifies the fitting procedure, because
there is no need to optimize with regards to the illumination parameters.
Furthermore, we propose a new fast algorithm for fitting the 3DMM in arbitrary
images. Finally, we have captured the first 3D facial database with relatively
unconstrained conditions and report quantitative evaluations with
state-of-the-art performance. Complementary qualitative reconstruction results
are demonstrated on standard "in-the-wild" facial databases. An open source
implementation of our technique is released as part of the Menpo Project.
| no_new_dataset | 0.945651 |
1701.05378 | Burak Civek | Burak C. Civek and Suleyman S. Kozat | Efficient Implementation Of Newton-Raphson Methods For Sequential Data
Prediction | null | null | null | null | cs.DS cs.CC cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the problem of sequential linear data prediction for real life
big data applications. The second order algorithms, i.e., Newton-Raphson
Methods, asymptotically achieve the performance of the "best" possible linear
data predictor much faster compared to the first order algorithms, e.g., Online
Gradient Descent. However, implementation of these methods is not usually
feasible in big data applications because of the extremely high computational
needs. Regular implementation of the Newton-Raphson Methods requires a
computational complexity in the order of $O(M^2)$ for an $M$ dimensional
feature vector, while the first order algorithms need only $O(M)$. To this end,
in order to eliminate this gap, we introduce a highly efficient implementation
reducing the computational complexity of the Newton-Raphson Methods from
quadratic to linear scale. The presented algorithm provides the well-known
merits of the second order methods while offering the computational complexity
of $O(M)$. We utilize the shifted nature of the consecutive feature vectors and
do not rely on any statistical assumptions. Therefore, both regular and fast
implementations achieve the same performance in the sense of mean square error.
We demonstrate the computational efficiency of our algorithm on real life
sequential big datasets. We also illustrate that the presented algorithm is
numerically stable.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 11:34:17 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Civek",
"Burak C.",
""
],
[
"Kozat",
"Suleyman S.",
""
]
] | TITLE: Efficient Implementation Of Newton-Raphson Methods For Sequential Data
Prediction
ABSTRACT: We investigate the problem of sequential linear data prediction for real life
big data applications. The second order algorithms, i.e., Newton-Raphson
Methods, asymptotically achieve the performance of the "best" possible linear
data predictor much faster compared to the first order algorithms, e.g., Online
Gradient Descent. However, implementation of these methods is not usually
feasible in big data applications because of the extremely high computational
needs. Regular implementation of the Newton-Raphson Methods requires a
computational complexity in the order of $O(M^2)$ for an $M$ dimensional
feature vector, while the first order algorithms need only $O(M)$. To this end,
in order to eliminate this gap, we introduce a highly efficient implementation
reducing the computational complexity of the Newton-Raphson Methods from
quadratic to linear scale. The presented algorithm provides the well-known
merits of the second order methods while offering the computational complexity
of $O(M)$. We utilize the shifted nature of the consecutive feature vectors and
do not rely on any statistical assumptions. Therefore, both regular and fast
implementations achieve the same performance in the sense of mean square error.
We demonstrate the computational efficiency of our algorithm on real life
sequential big datasets. We also illustrate that the presented algorithm is
numerically stable.
| no_new_dataset | 0.946498 |
1701.05432 | Anoop Cherian | Anoop Cherian, Piotr Koniusz, Stephen Gould | Higher-order Pooling of CNN Features via Kernel Linearization for Action
Recognition | 9 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most successful deep learning algorithms for action recognition extend models
designed for image-based tasks such as object recognition to video. Such
extensions are typically trained for actions on single video frames or very
short clips, and then their predictions from sliding-windows over the video
sequence are pooled for recognizing the action at the sequence level. Usually
this pooling step uses the first-order statistics of frame-level action
predictions. In this paper, we explore the advantages of using higher-order
correlations; specifically, we introduce Higher-order Kernel (HOK) descriptors
generated from the late fusion of CNN classifier scores from all the frames in
a sequence. To generate these descriptors, we use the idea of kernel
linearization. Specifically, a similarity kernel matrix, which captures the
temporal evolution of deep classifier scores, is first linearized into kernel
feature maps. The HOK descriptors are then generated from the higher-order
co-occurrences of these feature maps, and are then used as input to a
video-level classifier. We provide experiments on two fine-grained action
recognition datasets and show that our scheme leads to state-of-the-art
results.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 14:30:49 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Cherian",
"Anoop",
""
],
[
"Koniusz",
"Piotr",
""
],
[
"Gould",
"Stephen",
""
]
] | TITLE: Higher-order Pooling of CNN Features via Kernel Linearization for Action
Recognition
ABSTRACT: Most successful deep learning algorithms for action recognition extend models
designed for image-based tasks such as object recognition to video. Such
extensions are typically trained for actions on single video frames or very
short clips, and then their predictions from sliding-windows over the video
sequence are pooled for recognizing the action at the sequence level. Usually
this pooling step uses the first-order statistics of frame-level action
predictions. In this paper, we explore the advantages of using higher-order
correlations; specifically, we introduce Higher-order Kernel (HOK) descriptors
generated from the late fusion of CNN classifier scores from all the frames in
a sequence. To generate these descriptors, we use the idea of kernel
linearization. Specifically, a similarity kernel matrix, which captures the
temporal evolution of deep classifier scores, is first linearized into kernel
feature maps. The HOK descriptors are then generated from the higher-order
co-occurrences of these feature maps, and are then used as input to a
video-level classifier. We provide experiments on two fine-grained action
recognition datasets and show that our scheme leads to state-of-the-art
results.
| no_new_dataset | 0.95222 |
1701.05449 | Jerome Darmont | Varunya Attasena (ERIC), Nouria Harbi (ERIC), J\'er\^ome Darmont
(ERIC) | A Novel Multi-Secret Sharing Approach for Secure Data Warehousing and
On-Line Analysis Processing in the Cloud | null | International Journal of Data Warehousing and Mining, 11 (2),
pp.22 - 43 (2015) | 10.4018/ijdwm.2015040102 | null | cs.DB cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cloud computing helps reduce costs, increase business agility and deploy
solutions with a high return on investment for many types of applications,
including data warehouses and on-line analytical processing. However, storing
and transferring sensitive data into the cloud raises legitimate security
concerns. In this paper, we propose a new multi-secret sharing approach for
deploying data warehouses in the cloud and allowing on-line analysis
processing, while enforcing data privacy, integrity and availability. We first
validate the relevance of our approach theoretically and then experimentally
with both a simple random dataset and the Star Schema Benchmark. We also
demonstrate its superiority to related methods.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 14:54:21 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Attasena",
"Varunya",
"",
"ERIC"
],
[
"Harbi",
"Nouria",
"",
"ERIC"
],
[
"Darmont",
"Jérôme",
"",
"ERIC"
]
] | TITLE: A Novel Multi-Secret Sharing Approach for Secure Data Warehousing and
On-Line Analysis Processing in the Cloud
ABSTRACT: Cloud computing helps reduce costs, increase business agility and deploy
solutions with a high return on investment for many types of applications,
including data warehouses and on-line analytical processing. However, storing
and transferring sensitive data into the cloud raises legitimate security
concerns. In this paper, we propose a new multi-secret sharing approach for
deploying data warehouses in the cloud and allowing on-line analysis
processing, while enforcing data privacy, integrity and availability. We first
validate the relevance of our approach theoretically and then experimentally
with both a simple random dataset and the Star Schema Benchmark. We also
demonstrate its superiority to related methods.
| no_new_dataset | 0.949716 |
1701.05498 | Tomas Hodan | Tomas Hodan, Pavel Haluza, Stepan Obdrzalek, Jiri Matas, Manolis
Lourakis, Xenophon Zabulis | T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-less Objects | WACV 2017 | null | null | null | cs.CV cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce T-LESS, a new public dataset for estimating the 6D pose, i.e.
translation and rotation, of texture-less rigid objects. The dataset features
thirty industry-relevant objects with no significant texture and no
discriminative color or reflectance properties. The objects exhibit symmetries
and mutual similarities in shape and/or size. Compared to other datasets, a
unique property is that some of the objects are parts of others. The dataset
includes training and test images that were captured with three synchronized
sensors, specifically a structured-light and a time-of-flight RGB-D sensor and
a high-resolution RGB camera. There are approximately 39K training and 10K test
images from each sensor. Additionally, two types of 3D models are provided for
each object, i.e. a manually created CAD model and a semi-automatically
reconstructed one. Training images depict individual objects against a black
background. Test images originate from twenty test scenes having varying
complexity, which increases from simple scenes with several isolated objects to
very challenging ones with multiple instances of several objects and with a
high amount of clutter and occlusion. The images were captured from a
systematically sampled view sphere around the object/scene, and are annotated
with accurate ground truth 6D poses of all modeled objects. Initial evaluation
results indicate that the state of the art in 6D object pose estimation has
ample room for improvement, especially in difficult cases with significant
occlusion. The T-LESS dataset is available online at cmp.felk.cvut.cz/t-less.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2017 16:16:36 GMT"
}
] | 2017-01-20T00:00:00 | [
[
"Hodan",
"Tomas",
""
],
[
"Haluza",
"Pavel",
""
],
[
"Obdrzalek",
"Stepan",
""
],
[
"Matas",
"Jiri",
""
],
[
"Lourakis",
"Manolis",
""
],
[
"Zabulis",
"Xenophon",
""
]
] | TITLE: T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-less Objects
ABSTRACT: We introduce T-LESS, a new public dataset for estimating the 6D pose, i.e.
translation and rotation, of texture-less rigid objects. The dataset features
thirty industry-relevant objects with no significant texture and no
discriminative color or reflectance properties. The objects exhibit symmetries
and mutual similarities in shape and/or size. Compared to other datasets, a
unique property is that some of the objects are parts of others. The dataset
includes training and test images that were captured with three synchronized
sensors, specifically a structured-light and a time-of-flight RGB-D sensor and
a high-resolution RGB camera. There are approximately 39K training and 10K test
images from each sensor. Additionally, two types of 3D models are provided for
each object, i.e. a manually created CAD model and a semi-automatically
reconstructed one. Training images depict individual objects against a black
background. Test images originate from twenty test scenes having varying
complexity, which increases from simple scenes with several isolated objects to
very challenging ones with multiple instances of several objects and with a
high amount of clutter and occlusion. The images were captured from a
systematically sampled view sphere around the object/scene, and are annotated
with accurate ground truth 6D poses of all modeled objects. Initial evaluation
results indicate that the state of the art in 6D object pose estimation has
ample room for improvement, especially in difficult cases with significant
occlusion. The T-LESS dataset is available online at cmp.felk.cvut.cz/t-less.
| new_dataset | 0.968381 |
1604.05417 | Swami Sankaranarayanan | Swami Sankaranarayanan, Azadeh Alavi, Carlos Castillo, Rama Chellappa | Triplet Probabilistic Embedding for Face Verification and Clustering | Oral Paper in BTAS 2016; NVIDIA Best paper Award
(http://ieee-biometrics.org/btas2016/awards.html) | null | 10.1109/BTAS.2016.7791205 | null | cs.CV cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite significant progress made over the past twenty five years,
unconstrained face verification remains a challenging problem. This paper
proposes an approach that couples a deep CNN-based approach with a
low-dimensional discriminative embedding learned using triplet probability
constraints to solve the unconstrained face verification problem. Aside from
yielding performance improvements, this embedding provides significant
advantages in terms of memory and for post-processing operations like subject
specific clustering. Experiments on the challenging IJB-A dataset show that the
proposed algorithm performs comparably or better than the state of the art
methods in verification and identification metrics, while requiring much less
training data and training time. The superior performance of the proposed
method on the CFP dataset shows that the representation learned by our deep CNN
is robust to extreme pose variation. Furthermore, we demonstrate the robustness
of the deep features to challenges including age, pose, blur and clutter by
performing simple clustering experiments on both IJB-A and LFW datasets.
| [
{
"version": "v1",
"created": "Tue, 19 Apr 2016 03:29:56 GMT"
},
{
"version": "v2",
"created": "Sun, 8 May 2016 16:04:02 GMT"
},
{
"version": "v3",
"created": "Wed, 18 Jan 2017 03:10:44 GMT"
}
] | 2017-01-19T00:00:00 | [
[
"Sankaranarayanan",
"Swami",
""
],
[
"Alavi",
"Azadeh",
""
],
[
"Castillo",
"Carlos",
""
],
[
"Chellappa",
"Rama",
""
]
] | TITLE: Triplet Probabilistic Embedding for Face Verification and Clustering
ABSTRACT: Despite significant progress made over the past twenty five years,
unconstrained face verification remains a challenging problem. This paper
proposes an approach that couples a deep CNN-based approach with a
low-dimensional discriminative embedding learned using triplet probability
constraints to solve the unconstrained face verification problem. Aside from
yielding performance improvements, this embedding provides significant
advantages in terms of memory and for post-processing operations like subject
specific clustering. Experiments on the challenging IJB-A dataset show that the
proposed algorithm performs comparably or better than the state of the art
methods in verification and identification metrics, while requiring much less
training data and training time. The superior performance of the proposed
method on the CFP dataset shows that the representation learned by our deep CNN
is robust to extreme pose variation. Furthermore, we demonstrate the robustness
of the deep features to challenges including age, pose, blur and clutter by
performing simple clustering experiments on both IJB-A and LFW datasets.
| no_new_dataset | 0.949153 |
1701.04819 | Primoz Kajdic | P. Kajdic, O. Alexandrova, M. Maksimovic, C. Lacombe and A. N.
Fazakerley | Suprathermal electron strahl widths in the presence of narrow-band
whistler waves in the solar wind | Published in ApJ | Kajdic et al., TheApJ, 833, 172, 2016 | 10.3847/1538-4357/833/2/172 | null | physics.space-ph astro-ph.EP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We perform the first statistical study of the effects of the interaction of
suprathermal electrons with narrow-band whistler mode waves in the solar wind.
We show that this interaction does occur and that it is associated with
enhanced widths of the so called strahl component. The latter is directed along
the inter- planetary magnetic field away from the Sun. We do the study by
comparing the strahl pitch angle widths in the solar wind at 1AU in the absence
of large scale discontinuities and transient structures, such as interplanetary
shocks, interplanetary coronal mass ejections, stream interaction regions, etc.
during times when the whistler mode waves were present and when they were
absent. This is done by using the data from two Cluster instruments: STAFF data
in frequency range between ~0.1 Hz and ~200 Hz were used for determining the
wave properties and PEACE datasets at twelve central energies between ~57 eV
(equivalent to ~10 typical electron thermal energies in the solar wind, E_T )
and ~676 eV (~113 E_T ) for pitch angle measurements. Statistical analysis
shows that during the inter- vals with the whistler waves the strahl component
on average exhibits pitch angle widths between 2 and 12 degrees larger than
during the intervals when these waves are not present. The largest difference
is obtained for the electron central energy of ~344 eV (~57 E_T ).
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 17:12:01 GMT"
}
] | 2017-01-19T00:00:00 | [
[
"Kajdic",
"P.",
""
],
[
"Alexandrova",
"O.",
""
],
[
"Maksimovic",
"M.",
""
],
[
"Lacombe",
"C.",
""
],
[
"Fazakerley",
"A. N.",
""
]
] | TITLE: Suprathermal electron strahl widths in the presence of narrow-band
whistler waves in the solar wind
ABSTRACT: We perform the first statistical study of the effects of the interaction of
suprathermal electrons with narrow-band whistler mode waves in the solar wind.
We show that this interaction does occur and that it is associated with
enhanced widths of the so called strahl component. The latter is directed along
the inter- planetary magnetic field away from the Sun. We do the study by
comparing the strahl pitch angle widths in the solar wind at 1AU in the absence
of large scale discontinuities and transient structures, such as interplanetary
shocks, interplanetary coronal mass ejections, stream interaction regions, etc.
during times when the whistler mode waves were present and when they were
absent. This is done by using the data from two Cluster instruments: STAFF data
in frequency range between ~0.1 Hz and ~200 Hz were used for determining the
wave properties and PEACE datasets at twelve central energies between ~57 eV
(equivalent to ~10 typical electron thermal energies in the solar wind, E_T )
and ~676 eV (~113 E_T ) for pitch angle measurements. Statistical analysis
shows that during the inter- vals with the whistler waves the strahl component
on average exhibits pitch angle widths between 2 and 12 degrees larger than
during the intervals when these waves are not present. The largest difference
is obtained for the electron central energy of ~344 eV (~57 E_T ).
| no_new_dataset | 0.946892 |
1701.04934 | Swati Agarwal | Swati Agarwal and Ashish Sureka | Investigating the Application of Common-Sense Knowledge-Base for
Identifying Term Obfuscation in Adversarial Communication | This paper is an extended and detailed version of our (same authors)
previous paper (regular paper) published at COMSNETS2015 | S. Agarwal and A. Sureka, "Using common-sense knowledge-base for
detecting word obfuscation in adversarial communication," 2015 7th
International Conference on Communication Systems and Networks (COMSNETS),
Bangalore, 2015, pp. 1-6 | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Word obfuscation or substitution means replacing one word with another word
in a sentence to conceal the textual content or communication. Word obfuscation
is used in adversarial communication by terrorist or criminals for conveying
their messages without getting red-flagged by security and intelligence
agencies intercepting or scanning messages (such as emails and telephone
conversations). ConceptNet is a freely available semantic network represented
as a directed graph consisting of nodes as concepts and edges as assertions of
common sense about these concepts. We present a solution approach exploiting
vast amount of semantic knowledge in ConceptNet for addressing the technically
challenging problem of word substitution in adversarial communication. We frame
the given problem as a textual reasoning and context inference task and utilize
ConceptNet's natural-language-processing tool-kit for determining word
substitution. We use ConceptNet to compute the conceptual similarity between
any two given terms and define a Mean Average Conceptual Similarity (MACS)
metric to identify out-of-context terms. The test-bed to evaluate our proposed
approach consists of Enron email dataset (having over 600000 emails generated
by 158 employees of Enron Corporation) and Brown corpus (totaling about a
million words drawn from a wide variety of sources). We implement word
substitution techniques used by previous researches to generate a test dataset.
We conduct a series of experiments consisting of word substitution methods used
in the past to evaluate our approach. Experimental results reveal that the
proposed approach is effective.
| [
{
"version": "v1",
"created": "Wed, 18 Jan 2017 03:36:33 GMT"
}
] | 2017-01-19T00:00:00 | [
[
"Agarwal",
"Swati",
""
],
[
"Sureka",
"Ashish",
""
]
] | TITLE: Investigating the Application of Common-Sense Knowledge-Base for
Identifying Term Obfuscation in Adversarial Communication
ABSTRACT: Word obfuscation or substitution means replacing one word with another word
in a sentence to conceal the textual content or communication. Word obfuscation
is used in adversarial communication by terrorist or criminals for conveying
their messages without getting red-flagged by security and intelligence
agencies intercepting or scanning messages (such as emails and telephone
conversations). ConceptNet is a freely available semantic network represented
as a directed graph consisting of nodes as concepts and edges as assertions of
common sense about these concepts. We present a solution approach exploiting
vast amount of semantic knowledge in ConceptNet for addressing the technically
challenging problem of word substitution in adversarial communication. We frame
the given problem as a textual reasoning and context inference task and utilize
ConceptNet's natural-language-processing tool-kit for determining word
substitution. We use ConceptNet to compute the conceptual similarity between
any two given terms and define a Mean Average Conceptual Similarity (MACS)
metric to identify out-of-context terms. The test-bed to evaluate our proposed
approach consists of Enron email dataset (having over 600000 emails generated
by 158 employees of Enron Corporation) and Brown corpus (totaling about a
million words drawn from a wide variety of sources). We implement word
substitution techniques used by previous researches to generate a test dataset.
We conduct a series of experiments consisting of word substitution methods used
in the past to evaluate our approach. Experimental results reveal that the
proposed approach is effective.
| new_dataset | 0.960175 |
1701.04949 | Volodymyr Turchenko | Volodymyr Turchenko, Eric Chalmers, Artur Luczak | A Deep Convolutional Auto-Encoder with Pooling - Unpooling Layers in
Caffe | 21 pages, 11 figures, 5 tables, 62 references | null | null | null | cs.NE cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents the development of several models of a deep convolutional
auto-encoder in the Caffe deep learning framework and their experimental
evaluation on the example of MNIST dataset. We have created five models of a
convolutional auto-encoder which differ architecturally by the presence or
absence of pooling and unpooling layers in the auto-encoder's encoder and
decoder parts. Our results show that the developed models provide very good
results in dimensionality reduction and unsupervised clustering tasks, and
small classification errors when we used the learned internal code as an input
of a supervised linear classifier and multi-layer perceptron. The best results
were provided by a model where the encoder part contains convolutional and
pooling layers, followed by an analogous decoder part with deconvolution and
unpooling layers without the use of switch variables in the decoder part. The
paper also discusses practical details of the creation of a deep convolutional
auto-encoder in the very popular Caffe deep learning framework. We believe that
our approach and results presented in this paper could help other researchers
to build efficient deep neural network architectures in the future.
| [
{
"version": "v1",
"created": "Wed, 18 Jan 2017 05:24:24 GMT"
}
] | 2017-01-19T00:00:00 | [
[
"Turchenko",
"Volodymyr",
""
],
[
"Chalmers",
"Eric",
""
],
[
"Luczak",
"Artur",
""
]
] | TITLE: A Deep Convolutional Auto-Encoder with Pooling - Unpooling Layers in
Caffe
ABSTRACT: This paper presents the development of several models of a deep convolutional
auto-encoder in the Caffe deep learning framework and their experimental
evaluation on the example of MNIST dataset. We have created five models of a
convolutional auto-encoder which differ architecturally by the presence or
absence of pooling and unpooling layers in the auto-encoder's encoder and
decoder parts. Our results show that the developed models provide very good
results in dimensionality reduction and unsupervised clustering tasks, and
small classification errors when we used the learned internal code as an input
of a supervised linear classifier and multi-layer perceptron. The best results
were provided by a model where the encoder part contains convolutional and
pooling layers, followed by an analogous decoder part with deconvolution and
unpooling layers without the use of switch variables in the decoder part. The
paper also discusses practical details of the creation of a deep convolutional
auto-encoder in the very popular Caffe deep learning framework. We believe that
our approach and results presented in this paper could help other researchers
to build efficient deep neural network architectures in the future.
| no_new_dataset | 0.950365 |
1701.05149 | G\"urkan Alpaslan | G\"urkan Alpaslan | Comparison of the Efficiency of Different Algorithms on Recommendation
System Design: a Case Study | null | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | By the growing trend of online shopping and e-commerce websites,
recommendation systems have gained more importance in recent years in order to
increase the sales ratios of companies. Different algorithms on recommendation
systems are used and every one produce different results. Every algorithm on
this area have positive and negative attributes. The purpose of the research is
to test the different algorithms for choosing the best one according as
structure of dataset and aims of developers. For this purpose, threshold and
k-means based collaborative filtering and content-based filtering algorithms
are utilized on the dataset contains 100*73421 matrix length. What are the
differences and effects of these different algorithms on the same dataset? What
are the challenges of the algorithms? What criteria are more important in order
to evaluate a recommendation systems? In the study, we answer these crucial
problems with the case study.
| [
{
"version": "v1",
"created": "Sun, 1 Jan 2017 17:58:38 GMT"
}
] | 2017-01-19T00:00:00 | [
[
"Alpaslan",
"Gürkan",
""
]
] | TITLE: Comparison of the Efficiency of Different Algorithms on Recommendation
System Design: a Case Study
ABSTRACT: By the growing trend of online shopping and e-commerce websites,
recommendation systems have gained more importance in recent years in order to
increase the sales ratios of companies. Different algorithms on recommendation
systems are used and every one produce different results. Every algorithm on
this area have positive and negative attributes. The purpose of the research is
to test the different algorithms for choosing the best one according as
structure of dataset and aims of developers. For this purpose, threshold and
k-means based collaborative filtering and content-based filtering algorithms
are utilized on the dataset contains 100*73421 matrix length. What are the
differences and effects of these different algorithms on the same dataset? What
are the challenges of the algorithms? What criteria are more important in order
to evaluate a recommendation systems? In the study, we answer these crucial
problems with the case study.
| no_new_dataset | 0.950641 |
1605.03733 | Riccardo Sven Risuleo | Riccardo Sven Risuleo and Giulio Bottegal and H{\aa}kan Hjalmarsson | Kernel-based system identification from noisy and incomplete
input-output data | 16 pages, submitted to IEEE Conference on Decision and Control 2016 | null | 10.1109/CDC.2016.7798567 | null | cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this contribution, we propose a kernel-based method for the identification
of linear systems from noisy and incomplete input-output datasets. We model the
impulse response of the system as a Gaussian process whose covariance matrix is
given by the recently introduced stable spline kernel. We adopt an empirical
Bayes approach to estimate the posterior distribution of the impulse response
given the data. The noiseless and missing data samples, together with the
kernel hyperparameters, are estimated maximizing the joint marginal likelihood
of the input and output measurements. To compute the marginal-likelihood
maximizer, we build a solution scheme based on the Expectation-Maximization
method. Simulations on a benchmark dataset show the effectiveness of the
method.
| [
{
"version": "v1",
"created": "Thu, 12 May 2016 09:04:23 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Risuleo",
"Riccardo Sven",
""
],
[
"Bottegal",
"Giulio",
""
],
[
"Hjalmarsson",
"Håkan",
""
]
] | TITLE: Kernel-based system identification from noisy and incomplete
input-output data
ABSTRACT: In this contribution, we propose a kernel-based method for the identification
of linear systems from noisy and incomplete input-output datasets. We model the
impulse response of the system as a Gaussian process whose covariance matrix is
given by the recently introduced stable spline kernel. We adopt an empirical
Bayes approach to estimate the posterior distribution of the impulse response
given the data. The noiseless and missing data samples, together with the
kernel hyperparameters, are estimated maximizing the joint marginal likelihood
of the input and output measurements. To compute the marginal-likelihood
maximizer, we build a solution scheme based on the Expectation-Maximization
method. Simulations on a benchmark dataset show the effectiveness of the
method.
| no_new_dataset | 0.947769 |
1606.00305 | Yang Li | Yang Li, Chunxiao Fan, Yong Li, Qiong Wu, Yue Ming | Improving Deep Neural Network with Multiple Parametric Exponential
Linear Units | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Activation function is crucial to the recent successes of deep neural
networks. In this paper, we first propose a new activation function, Multiple
Parametric Exponential Linear Units (MPELU), aiming to generalize and unify the
rectified and exponential linear units. As the generalized form, MPELU shares
the advantages of Parametric Rectified Linear Unit (PReLU) and Exponential
Linear Unit (ELU), leading to better classification performance and convergence
property. In addition, weight initialization is very important to train very
deep networks. The existing methods laid a solid foundation for networks using
rectified linear units but not for exponential linear units. This paper
complements the current theory and extends it to the wider range. Specifically,
we put forward a way of initialization, enabling training of very deep networks
using exponential linear units. Experiments demonstrate that the proposed
initialization not only helps the training process but leads to better
generalization performance. Finally, utilizing the proposed activation function
and initialization, we present a deep MPELU residual architecture that achieves
state-of-the-art performance on the CIFAR-10/100 datasets. The code is
available at https://github.com/Coldmooon/Code-for-MPELU.
| [
{
"version": "v1",
"created": "Wed, 1 Jun 2016 14:33:17 GMT"
},
{
"version": "v2",
"created": "Tue, 9 Aug 2016 10:10:41 GMT"
},
{
"version": "v3",
"created": "Tue, 17 Jan 2017 08:44:56 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Li",
"Yang",
""
],
[
"Fan",
"Chunxiao",
""
],
[
"Li",
"Yong",
""
],
[
"Wu",
"Qiong",
""
],
[
"Ming",
"Yue",
""
]
] | TITLE: Improving Deep Neural Network with Multiple Parametric Exponential
Linear Units
ABSTRACT: Activation function is crucial to the recent successes of deep neural
networks. In this paper, we first propose a new activation function, Multiple
Parametric Exponential Linear Units (MPELU), aiming to generalize and unify the
rectified and exponential linear units. As the generalized form, MPELU shares
the advantages of Parametric Rectified Linear Unit (PReLU) and Exponential
Linear Unit (ELU), leading to better classification performance and convergence
property. In addition, weight initialization is very important to train very
deep networks. The existing methods laid a solid foundation for networks using
rectified linear units but not for exponential linear units. This paper
complements the current theory and extends it to the wider range. Specifically,
we put forward a way of initialization, enabling training of very deep networks
using exponential linear units. Experiments demonstrate that the proposed
initialization not only helps the training process but leads to better
generalization performance. Finally, utilizing the proposed activation function
and initialization, we present a deep MPELU residual architecture that achieves
state-of-the-art performance on the CIFAR-10/100 datasets. The code is
available at https://github.com/Coldmooon/Code-for-MPELU.
| no_new_dataset | 0.952086 |
1610.05949 | Raul Mur-Artal | Raul Mur-Artal and Juan D. Tardos | Visual-Inertial Monocular SLAM with Map Reuse | Accepted for publication in IEEE Robotics and Automation Letters | null | 10.1109/LRA.2017.2653359 | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years there have been excellent results in Visual-Inertial Odometry
techniques, which aim to compute the incremental motion of the sensor with high
accuracy and robustness. However these approaches lack the capability to close
loops, and trajectory estimation accumulates drift even if the sensor is
continually revisiting the same place. In this work we present a novel
tightly-coupled Visual-Inertial Simultaneous Localization and Mapping system
that is able to close loops and reuse its map to achieve zero-drift
localization in already mapped areas. While our approach can be applied to any
camera configuration, we address here the most general problem of a monocular
camera, with its well-known scale ambiguity. We also propose a novel IMU
initialization method, which computes the scale, the gravity direction, the
velocity, and gyroscope and accelerometer biases, in a few seconds with high
accuracy. We test our system in the 11 sequences of a recent micro-aerial
vehicle public dataset achieving a typical scale factor error of 1% and
centimeter precision. We compare to the state-of-the-art in visual-inertial
odometry in sequences with revisiting, proving the better accuracy of our
method due to map reuse and no drift accumulation.
| [
{
"version": "v1",
"created": "Wed, 19 Oct 2016 10:17:16 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jan 2017 15:45:14 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Mur-Artal",
"Raul",
""
],
[
"Tardos",
"Juan D.",
""
]
] | TITLE: Visual-Inertial Monocular SLAM with Map Reuse
ABSTRACT: In recent years there have been excellent results in Visual-Inertial Odometry
techniques, which aim to compute the incremental motion of the sensor with high
accuracy and robustness. However these approaches lack the capability to close
loops, and trajectory estimation accumulates drift even if the sensor is
continually revisiting the same place. In this work we present a novel
tightly-coupled Visual-Inertial Simultaneous Localization and Mapping system
that is able to close loops and reuse its map to achieve zero-drift
localization in already mapped areas. While our approach can be applied to any
camera configuration, we address here the most general problem of a monocular
camera, with its well-known scale ambiguity. We also propose a novel IMU
initialization method, which computes the scale, the gravity direction, the
velocity, and gyroscope and accelerometer biases, in a few seconds with high
accuracy. We test our system in the 11 sequences of a recent micro-aerial
vehicle public dataset achieving a typical scale factor error of 1% and
centimeter precision. We compare to the state-of-the-art in visual-inertial
odometry in sequences with revisiting, proving the better accuracy of our
method due to map reuse and no drift accumulation.
| no_new_dataset | 0.948298 |
1611.00303 | Timothy O'Shea | Timothy J. O'Shea, Nathan West, Matthew Vondal, T. Charles Clancy | Semi-Supervised Radio Signal Identification | null | null | null | null | cs.LG cs.IT math.IT stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radio emitter recognition in dense multi-user environments is an important
tool for optimizing spectrum utilization, identifying and minimizing
interference, and enforcing spectrum policy. Radio data is readily available
and easy to obtain from an antenna, but labeled and curated data is often
scarce making supervised learning strategies difficult and time consuming in
practice. We demonstrate that semi-supervised learning techniques can be used
to scale learning beyond supervised datasets, allowing for discerning and
recalling new radio signals by using sparse signal representations based on
both unsupervised and supervised methods for nonlinear feature learning and
clustering methods.
| [
{
"version": "v1",
"created": "Tue, 1 Nov 2016 17:21:50 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jan 2017 18:23:49 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"O'Shea",
"Timothy J.",
""
],
[
"West",
"Nathan",
""
],
[
"Vondal",
"Matthew",
""
],
[
"Clancy",
"T. Charles",
""
]
] | TITLE: Semi-Supervised Radio Signal Identification
ABSTRACT: Radio emitter recognition in dense multi-user environments is an important
tool for optimizing spectrum utilization, identifying and minimizing
interference, and enforcing spectrum policy. Radio data is readily available
and easy to obtain from an antenna, but labeled and curated data is often
scarce making supervised learning strategies difficult and time consuming in
practice. We demonstrate that semi-supervised learning techniques can be used
to scale learning beyond supervised datasets, allowing for discerning and
recalling new radio signals by using sparse signal representations based on
both unsupervised and supervised methods for nonlinear feature learning and
clustering methods.
| no_new_dataset | 0.949059 |
1612.00220 | Mark Marsden | Mark Marsden, Kevin McGuinness, Suzanne Little, Noel E. O'Connor | Fully Convolutional Crowd Counting On Highly Congested Scenes | 7 pages , VISAPP 2017 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we advance the state-of-the-art for crowd counting in high
density scenes by further exploring the idea of a fully convolutional crowd
counting model introduced by (Zhang et al., 2016). Producing an accurate and
robust crowd count estimator using computer vision techniques has attracted
significant research interest in recent years. Applications for crowd counting
systems exist in many diverse areas including city planning, retail, and of
course general public safety. Developing a highly generalised counting model
that can be deployed in any surveillance scenario with any camera perspective
is the key objective for research in this area. Techniques developed in the
past have generally performed poorly in highly congested scenes with several
thousands of people in frame (Rodriguez et al., 2011). Our approach, influenced
by the work of (Zhang et al., 2016), consists of the following contributions:
(1) A training set augmentation scheme that minimises redundancy among training
samples to improve model generalisation and overall counting performance; (2) a
deep, single column, fully convolutional network (FCN) architecture; (3) a
multi-scale averaging step during inference. The developed technique can
analyse images of any resolution or aspect ratio and achieves state-of-the-art
counting performance on the Shanghaitech Part B and UCF CC 50 datasets as well
as competitive performance on Shanghaitech Part A.
| [
{
"version": "v1",
"created": "Thu, 1 Dec 2016 12:24:35 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jan 2017 15:00:46 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Marsden",
"Mark",
""
],
[
"McGuinness",
"Kevin",
""
],
[
"Little",
"Suzanne",
""
],
[
"O'Connor",
"Noel E.",
""
]
] | TITLE: Fully Convolutional Crowd Counting On Highly Congested Scenes
ABSTRACT: In this paper we advance the state-of-the-art for crowd counting in high
density scenes by further exploring the idea of a fully convolutional crowd
counting model introduced by (Zhang et al., 2016). Producing an accurate and
robust crowd count estimator using computer vision techniques has attracted
significant research interest in recent years. Applications for crowd counting
systems exist in many diverse areas including city planning, retail, and of
course general public safety. Developing a highly generalised counting model
that can be deployed in any surveillance scenario with any camera perspective
is the key objective for research in this area. Techniques developed in the
past have generally performed poorly in highly congested scenes with several
thousands of people in frame (Rodriguez et al., 2011). Our approach, influenced
by the work of (Zhang et al., 2016), consists of the following contributions:
(1) A training set augmentation scheme that minimises redundancy among training
samples to improve model generalisation and overall counting performance; (2) a
deep, single column, fully convolutional network (FCN) architecture; (3) a
multi-scale averaging step during inference. The developed technique can
analyse images of any resolution or aspect ratio and achieves state-of-the-art
counting performance on the Shanghaitech Part B and UCF CC 50 datasets as well
as competitive performance on Shanghaitech Part A.
| no_new_dataset | 0.952131 |
1701.04568 | Mahesh Gorijala | Mahesh Gorijala, Ambedkar Dukkipati | Image Generation and Editing with Variational Info Generative
AdversarialNetworks | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently there has been an enormous interest in generative models for images
in deep learning. In pursuit of this, Generative Adversarial Networks (GAN) and
Variational Auto-Encoder (VAE) have surfaced as two most prominent and popular
models. While VAEs tend to produce excellent reconstructions but blurry
samples, GANs generate sharp but slightly distorted images. In this paper we
propose a new model called Variational InfoGAN (ViGAN). Our aim is two fold:
(i) To generated new images conditioned on visual descriptions, and (ii) modify
the image, by fixing the latent representation of image and varying the visual
description. We evaluate our model on Labeled Faces in the Wild (LFW), celebA
and a modified version of MNIST datasets and demonstrate the ability of our
model to generate new images as well as to modify a given image by changing
attributes.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 08:48:28 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Gorijala",
"Mahesh",
""
],
[
"Dukkipati",
"Ambedkar",
""
]
] | TITLE: Image Generation and Editing with Variational Info Generative
AdversarialNetworks
ABSTRACT: Recently there has been an enormous interest in generative models for images
in deep learning. In pursuit of this, Generative Adversarial Networks (GAN) and
Variational Auto-Encoder (VAE) have surfaced as two most prominent and popular
models. While VAEs tend to produce excellent reconstructions but blurry
samples, GANs generate sharp but slightly distorted images. In this paper we
propose a new model called Variational InfoGAN (ViGAN). Our aim is two fold:
(i) To generated new images conditioned on visual descriptions, and (ii) modify
the image, by fixing the latent representation of image and varying the visual
description. We evaluate our model on Labeled Faces in the Wild (LFW), celebA
and a modified version of MNIST datasets and demonstrate the ability of our
model to generate new images as well as to modify a given image by changing
attributes.
| no_new_dataset | 0.951369 |
1701.04600 | Amit Awekar | Siddhesh Khandelwal, Amit Awekar | Faster K-Means Cluster Estimation | 6 pages, Accepted at ECIR 2017 | null | null | null | cs.LG cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There has been considerable work on improving popular clustering algorithm
`K-means' in terms of mean squared error (MSE) and speed, both. However, most
of the k-means variants tend to compute distance of each data point to each
cluster centroid for every iteration. We propose a fast heuristic to overcome
this bottleneck with only marginal increase in MSE. We observe that across all
iterations of K-means, a data point changes its membership only among a small
subset of clusters. Our heuristic predicts such clusters for each data point by
looking at nearby clusters after the first iteration of k-means. We augment
well known variants of k-means with our heuristic to demonstrate effectiveness
of our heuristic. For various synthetic and real-world datasets, our heuristic
achieves speed-up of up-to 3 times when compared to efficient variants of
k-means.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 10:00:51 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Khandelwal",
"Siddhesh",
""
],
[
"Awekar",
"Amit",
""
]
] | TITLE: Faster K-Means Cluster Estimation
ABSTRACT: There has been considerable work on improving popular clustering algorithm
`K-means' in terms of mean squared error (MSE) and speed, both. However, most
of the k-means variants tend to compute distance of each data point to each
cluster centroid for every iteration. We propose a fast heuristic to overcome
this bottleneck with only marginal increase in MSE. We observe that across all
iterations of K-means, a data point changes its membership only among a small
subset of clusters. Our heuristic predicts such clusters for each data point by
looking at nearby clusters after the first iteration of k-means. We augment
well known variants of k-means with our heuristic to demonstrate effectiveness
of our heuristic. For various synthetic and real-world datasets, our heuristic
achieves speed-up of up-to 3 times when compared to efficient variants of
k-means.
| no_new_dataset | 0.945901 |
1701.04653 | Marzieh Saeidi Marzieh Saeidi | Marzieh Saeidi, Alessandro Venerandi, Licia Capra and Sebastian Riedel | Community Question Answering Platforms vs. Twitter for Predicting
Characteristics of Urban Neighbourhoods | Submitted to ICWSM2017 | null | null | null | cs.CL cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we investigate whether text from a Community Question
Answering (QA) platform can be used to predict and describe real-world
attributes. We experiment with predicting a wide range of 62 demographic
attributes for neighbourhoods of London. We use the text from QA platform of
Yahoo! Answers and compare our results to the ones obtained from Twitter
microblogs. Outcomes show that the correlation between the predicted
demographic attributes using text from Yahoo! Answers discussions and the
observed demographic attributes can reach an average Pearson correlation
coefficient of \r{ho} = 0.54, slightly higher than the predictions obtained
using Twitter data. Our qualitative analysis indicates that there is semantic
relatedness between the highest correlated terms extracted from both datasets
and their relative demographic attributes. Furthermore, the correlations
highlight the different natures of the information contained in Yahoo! Answers
and Twitter. While the former seems to offer a more encyclopedic content, the
latter provides information related to the current sociocultural aspects or
phenomena.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 12:53:19 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Saeidi",
"Marzieh",
""
],
[
"Venerandi",
"Alessandro",
""
],
[
"Capra",
"Licia",
""
],
[
"Riedel",
"Sebastian",
""
]
] | TITLE: Community Question Answering Platforms vs. Twitter for Predicting
Characteristics of Urban Neighbourhoods
ABSTRACT: In this paper, we investigate whether text from a Community Question
Answering (QA) platform can be used to predict and describe real-world
attributes. We experiment with predicting a wide range of 62 demographic
attributes for neighbourhoods of London. We use the text from QA platform of
Yahoo! Answers and compare our results to the ones obtained from Twitter
microblogs. Outcomes show that the correlation between the predicted
demographic attributes using text from Yahoo! Answers discussions and the
observed demographic attributes can reach an average Pearson correlation
coefficient of \r{ho} = 0.54, slightly higher than the predictions obtained
using Twitter data. Our qualitative analysis indicates that there is semantic
relatedness between the highest correlated terms extracted from both datasets
and their relative demographic attributes. Furthermore, the correlations
highlight the different natures of the information contained in Yahoo! Answers
and Twitter. While the former seems to offer a more encyclopedic content, the
latter provides information related to the current sociocultural aspects or
phenomena.
| no_new_dataset | 0.953101 |
1701.04693 | Sepehr Valipour | Sepehr Valipour, Camilo Perez, Martin Jagersand | Incremental Learning for Robot Perception through HRI | null | null | null | null | cs.RO cs.HC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scene understanding and object recognition is a difficult to achieve yet
crucial skill for robots. Recently, Convolutional Neural Networks (CNN), have
shown success in this task. However, there is still a gap between their
performance on image datasets and real-world robotics scenarios. We present a
novel paradigm for incrementally improving a robot's visual perception through
active human interaction. In this paradigm, the user introduces novel objects
to the robot by means of pointing and voice commands. Given this information,
the robot visually explores the object and adds images from it to re-train the
perception module. Our base perception module is based on recent development in
object detection and recognition using deep learning. Our method leverages
state of the art CNNs from off-line batch learning, human guidance, robot
exploration and incremental on-line learning.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 14:29:05 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Valipour",
"Sepehr",
""
],
[
"Perez",
"Camilo",
""
],
[
"Jagersand",
"Martin",
""
]
] | TITLE: Incremental Learning for Robot Perception through HRI
ABSTRACT: Scene understanding and object recognition is a difficult to achieve yet
crucial skill for robots. Recently, Convolutional Neural Networks (CNN), have
shown success in this task. However, there is still a gap between their
performance on image datasets and real-world robotics scenarios. We present a
novel paradigm for incrementally improving a robot's visual perception through
active human interaction. In this paradigm, the user introduces novel objects
to the robot by means of pointing and voice commands. Given this information,
the robot visually explores the object and adds images from it to re-train the
perception module. Our base perception module is based on recent development in
object detection and recognition using deep learning. Our method leverages
state of the art CNNs from off-line batch learning, human guidance, robot
exploration and incremental on-line learning.
| no_new_dataset | 0.950273 |
1701.04769 | Unaiza Ahsan | Unaiza Ahsan, Chen Sun, James Hays and Irfan Essa | Complex Event Recognition from Images with Few Training Examples | Accepted to Winter Applications of Computer Vision (WACV'17) | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We propose to leverage concept-level representations for complex event
recognition in photographs given limited training examples. We introduce a
novel framework to discover event concept attributes from the web and use that
to extract semantic features from images and classify them into social event
categories with few training examples. Discovered concepts include a variety of
objects, scenes, actions and event sub-types, leading to a discriminative and
compact representation for event images. Web images are obtained for each
discovered event concept and we use (pretrained) CNN features to train concept
classifiers. Extensive experiments on challenging event datasets demonstrate
that our proposed method outperforms several baselines using deep CNN features
directly in classifying images into events with limited training examples. We
also demonstrate that our method achieves the best overall accuracy on a
dataset with unseen event categories using a single training example.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 17:16:55 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Ahsan",
"Unaiza",
""
],
[
"Sun",
"Chen",
""
],
[
"Hays",
"James",
""
],
[
"Essa",
"Irfan",
""
]
] | TITLE: Complex Event Recognition from Images with Few Training Examples
ABSTRACT: We propose to leverage concept-level representations for complex event
recognition in photographs given limited training examples. We introduce a
novel framework to discover event concept attributes from the web and use that
to extract semantic features from images and classify them into social event
categories with few training examples. Discovered concepts include a variety of
objects, scenes, actions and event sub-types, leading to a discriminative and
compact representation for event images. Web images are obtained for each
discovered event concept and we use (pretrained) CNN features to train concept
classifiers. Extensive experiments on challenging event datasets demonstrate
that our proposed method outperforms several baselines using deep CNN features
directly in classifying images into events with limited training examples. We
also demonstrate that our method achieves the best overall accuracy on a
dataset with unseen event categories using a single training example.
| no_new_dataset | 0.950915 |
1701.04783 | Lei Zheng | Lei Zheng, Vahid Noroozi, Philip S. Yu | Joint Deep Modeling of Users and Items Using Reviews for Recommendation | WSDM 2017 | null | null | null | cs.LG cs.IR | http://creativecommons.org/licenses/by/4.0/ | A large amount of information exists in reviews written by users. This source
of information has been ignored by most of the current recommender systems
while it can potentially alleviate the sparsity problem and improve the quality
of recommendations. In this paper, we present a deep model to learn item
properties and user behaviors jointly from review text. The proposed model,
named Deep Cooperative Neural Networks (DeepCoNN), consists of two parallel
neural networks coupled in the last layers. One of the networks focuses on
learning user behaviors exploiting reviews written by the user, and the other
one learns item properties from the reviews written for the item. A shared
layer is introduced on the top to couple these two networks together. The
shared layer enables latent factors learned for users and items to interact
with each other in a manner similar to factorization machine techniques.
Experimental results demonstrate that DeepCoNN significantly outperforms all
baseline recommender systems on a variety of datasets.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2017 17:46:04 GMT"
}
] | 2017-01-18T00:00:00 | [
[
"Zheng",
"Lei",
""
],
[
"Noroozi",
"Vahid",
""
],
[
"Yu",
"Philip S.",
""
]
] | TITLE: Joint Deep Modeling of Users and Items Using Reviews for Recommendation
ABSTRACT: A large amount of information exists in reviews written by users. This source
of information has been ignored by most of the current recommender systems
while it can potentially alleviate the sparsity problem and improve the quality
of recommendations. In this paper, we present a deep model to learn item
properties and user behaviors jointly from review text. The proposed model,
named Deep Cooperative Neural Networks (DeepCoNN), consists of two parallel
neural networks coupled in the last layers. One of the networks focuses on
learning user behaviors exploiting reviews written by the user, and the other
one learns item properties from the reviews written for the item. A shared
layer is introduced on the top to couple these two networks together. The
shared layer enables latent factors learned for users and items to interact
with each other in a manner similar to factorization machine techniques.
Experimental results demonstrate that DeepCoNN significantly outperforms all
baseline recommender systems on a variety of datasets.
| no_new_dataset | 0.950732 |
1606.00511 | Xi He | Xi He and Dheevatsa Mudigere and Mikhail Smelyanskiy and Martin
Tak\'a\v{c} | Distributed Hessian-Free Optimization for Deep Neural Network | null | null | null | null | cs.LG cs.DC math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Training deep neural network is a high dimensional and a highly non-convex
optimization problem. Stochastic gradient descent (SGD) algorithm and it's
variations are the current state-of-the-art solvers for this task. However, due
to non-covexity nature of the problem, it was observed that SGD slows down near
saddle point. Recent empirical work claim that by detecting and escaping saddle
point efficiently, it's more likely to improve training performance. With this
objective, we revisit Hessian-free optimization method for deep networks. We
also develop its distributed variant and demonstrate superior scaling potential
to SGD, which allows more efficiently utilizing larger computing resources thus
enabling large models and faster time to obtain desired solution. Furthermore,
unlike truncated Newton method (Marten's HF) that ignores negative curvature
information by using na\"ive conjugate gradient method and Gauss-Newton Hessian
approximation information - we propose a novel algorithm to explore negative
curvature direction by solving the sub-problem with stabilized bi-conjugate
method involving possible indefinite stochastic Hessian information. We show
that these techniques accelerate the training process for both the standard
MNIST dataset and also the TIMIT speech recognition problem, demonstrating
robust performance with upto an order of magnitude larger batch sizes. This
increased scaling potential is illustrated with near linear speed-up on upto 16
CPU nodes for a simple 4-layer network.
| [
{
"version": "v1",
"created": "Thu, 2 Jun 2016 00:39:03 GMT"
},
{
"version": "v2",
"created": "Sun, 15 Jan 2017 13:51:26 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"He",
"Xi",
""
],
[
"Mudigere",
"Dheevatsa",
""
],
[
"Smelyanskiy",
"Mikhail",
""
],
[
"Takáč",
"Martin",
""
]
] | TITLE: Distributed Hessian-Free Optimization for Deep Neural Network
ABSTRACT: Training deep neural network is a high dimensional and a highly non-convex
optimization problem. Stochastic gradient descent (SGD) algorithm and it's
variations are the current state-of-the-art solvers for this task. However, due
to non-covexity nature of the problem, it was observed that SGD slows down near
saddle point. Recent empirical work claim that by detecting and escaping saddle
point efficiently, it's more likely to improve training performance. With this
objective, we revisit Hessian-free optimization method for deep networks. We
also develop its distributed variant and demonstrate superior scaling potential
to SGD, which allows more efficiently utilizing larger computing resources thus
enabling large models and faster time to obtain desired solution. Furthermore,
unlike truncated Newton method (Marten's HF) that ignores negative curvature
information by using na\"ive conjugate gradient method and Gauss-Newton Hessian
approximation information - we propose a novel algorithm to explore negative
curvature direction by solving the sub-problem with stabilized bi-conjugate
method involving possible indefinite stochastic Hessian information. We show
that these techniques accelerate the training process for both the standard
MNIST dataset and also the TIMIT speech recognition problem, demonstrating
robust performance with upto an order of magnitude larger batch sizes. This
increased scaling potential is illustrated with near linear speed-up on upto 16
CPU nodes for a simple 4-layer network.
| no_new_dataset | 0.945851 |
1608.00486 | Conrad Sanderson | ZongYuan Ge, Chris McCool, Conrad Sanderson, Peng Wang, Lingqiao Liu,
Ian Reid, Peter Corke | Exploiting Temporal Information for DCNN-based Fine-Grained Object
Classification | International Conference on Digital Image Computing: Techniques and
Applications, 2016 | null | 10.1109/DICTA.2016.7797039 | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fine-grained classification is a relatively new field that has concentrated
on using information from a single image, while ignoring the enormous potential
of using video data to improve classification. In this work we present the
novel task of video-based fine-grained object classification, propose a
corresponding new video dataset, and perform a systematic study of several
recent deep convolutional neural network (DCNN) based approaches, which we
specifically adapt to the task. We evaluate three-dimensional DCNNs, two-stream
DCNNs, and bilinear DCNNs. Two forms of the two-stream approach are used, where
spatial and temporal data from two independent DCNNs are fused either via early
fusion (combination of the fully-connected layers) and late fusion
(concatenation of the softmax outputs of the DCNNs). For bilinear DCNNs,
information from the convolutional layers of the spatial and temporal DCNNs is
combined via local co-occurrences. We then fuse the bilinear DCNN and early
fusion of the two-stream approach to combine the spatial and temporal
information at the local and global level (Spatio-Temporal Co-occurrence).
Using the new and challenging video dataset of birds, classification
performance is improved from 23.1% (using single images) to 41.1% when using
the Spatio-Temporal Co-occurrence system. Incorporating automatically detected
bounding box location further improves the classification accuracy to 53.6%.
| [
{
"version": "v1",
"created": "Mon, 1 Aug 2016 16:34:16 GMT"
},
{
"version": "v2",
"created": "Wed, 31 Aug 2016 01:23:47 GMT"
},
{
"version": "v3",
"created": "Mon, 24 Oct 2016 06:40:02 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Ge",
"ZongYuan",
""
],
[
"McCool",
"Chris",
""
],
[
"Sanderson",
"Conrad",
""
],
[
"Wang",
"Peng",
""
],
[
"Liu",
"Lingqiao",
""
],
[
"Reid",
"Ian",
""
],
[
"Corke",
"Peter",
""
]
] | TITLE: Exploiting Temporal Information for DCNN-based Fine-Grained Object
Classification
ABSTRACT: Fine-grained classification is a relatively new field that has concentrated
on using information from a single image, while ignoring the enormous potential
of using video data to improve classification. In this work we present the
novel task of video-based fine-grained object classification, propose a
corresponding new video dataset, and perform a systematic study of several
recent deep convolutional neural network (DCNN) based approaches, which we
specifically adapt to the task. We evaluate three-dimensional DCNNs, two-stream
DCNNs, and bilinear DCNNs. Two forms of the two-stream approach are used, where
spatial and temporal data from two independent DCNNs are fused either via early
fusion (combination of the fully-connected layers) and late fusion
(concatenation of the softmax outputs of the DCNNs). For bilinear DCNNs,
information from the convolutional layers of the spatial and temporal DCNNs is
combined via local co-occurrences. We then fuse the bilinear DCNN and early
fusion of the two-stream approach to combine the spatial and temporal
information at the local and global level (Spatio-Temporal Co-occurrence).
Using the new and challenging video dataset of birds, classification
performance is improved from 23.1% (using single images) to 41.1% when using
the Spatio-Temporal Co-occurrence system. Incorporating automatically detected
bounding box location further improves the classification accuracy to 53.6%.
| no_new_dataset | 0.952618 |
1608.06770 | Emanuele Sansone | E. Sansone, K. Apostolidis, N. Conci, G. Boato, V. Mezaris, F.G.B. De
Natale | Automatic Synchronization of Multi-User Photo Galleries | ACCEPTED to IEEE Transactions on Multimedia | null | null | null | cs.MM cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we address the issue of photo galleries synchronization, where
pictures related to the same event are collected by different users. Existing
solutions to address the problem are usually based on unrealistic assumptions,
like time consistency across photo galleries, and often heavily rely on
heuristics, limiting therefore the applicability to real-world scenarios. We
propose a solution that achieves better generalization performance for the
synchronization task compared to the available literature. The method is
characterized by three stages: at first, deep convolutional neural network
features are used to assess the visual similarity among the photos; then, pairs
of similar photos are detected across different galleries and used to construct
a graph; eventually, a probabilistic graphical model is used to estimate the
temporal offset of each pair of galleries, by traversing the minimum spanning
tree extracted from this graph. The experimental evaluation is conducted on
four publicly available datasets covering different types of events,
demonstrating the strength of our proposed method. A thorough discussion of the
obtained results is provided for a critical assessment of the quality in
synchronization.
| [
{
"version": "v1",
"created": "Wed, 24 Aug 2016 10:17:16 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jan 2017 11:19:49 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Sansone",
"E.",
""
],
[
"Apostolidis",
"K.",
""
],
[
"Conci",
"N.",
""
],
[
"Boato",
"G.",
""
],
[
"Mezaris",
"V.",
""
],
[
"De Natale",
"F. G. B.",
""
]
] | TITLE: Automatic Synchronization of Multi-User Photo Galleries
ABSTRACT: In this paper we address the issue of photo galleries synchronization, where
pictures related to the same event are collected by different users. Existing
solutions to address the problem are usually based on unrealistic assumptions,
like time consistency across photo galleries, and often heavily rely on
heuristics, limiting therefore the applicability to real-world scenarios. We
propose a solution that achieves better generalization performance for the
synchronization task compared to the available literature. The method is
characterized by three stages: at first, deep convolutional neural network
features are used to assess the visual similarity among the photos; then, pairs
of similar photos are detected across different galleries and used to construct
a graph; eventually, a probabilistic graphical model is used to estimate the
temporal offset of each pair of galleries, by traversing the minimum spanning
tree extracted from this graph. The experimental evaluation is conducted on
four publicly available datasets covering different types of events,
demonstrating the strength of our proposed method. A thorough discussion of the
obtained results is provided for a critical assessment of the quality in
synchronization.
| no_new_dataset | 0.945248 |
1610.09091 | Shijia E | Shijia E, Yang Xiang, Mohan Zhang | Representation Learning Models for Entity Search | This paper has been withdrawn by the author because the proposed
model need to be re-evaluate | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We focus on the problem of learning distributed representations for entity
search queries, named entities, and their short descriptions. With our
representation learning models, the entity search query, named entity and
description can be represented as low-dimensional vectors. Our goal is to
develop a simple but effective model that can make the distributed
representations of query related entities similar to the query in the vector
space. Hence, we propose three kinds of learning strategies, and the difference
between them mainly lies in how to deal with the relationship between an entity
and its description. We analyze the strengths and weaknesses of each learning
strategy and validate our methods on public datasets which contain four kinds
of named entities, i.e., movies, TV shows, restaurants and celebrities. The
experimental results indicate that our proposed methods can adapt to different
types of entity search queries, and outperform the current state-of-the-art
methods based on keyword matching and vanilla word2vec models. Besides, the
proposed methods can be trained fast and be easily extended to other similar
tasks.
| [
{
"version": "v1",
"created": "Fri, 28 Oct 2016 06:33:33 GMT"
},
{
"version": "v2",
"created": "Tue, 20 Dec 2016 02:19:01 GMT"
},
{
"version": "v3",
"created": "Sun, 15 Jan 2017 13:57:23 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"E",
"Shijia",
""
],
[
"Xiang",
"Yang",
""
],
[
"Zhang",
"Mohan",
""
]
] | TITLE: Representation Learning Models for Entity Search
ABSTRACT: We focus on the problem of learning distributed representations for entity
search queries, named entities, and their short descriptions. With our
representation learning models, the entity search query, named entity and
description can be represented as low-dimensional vectors. Our goal is to
develop a simple but effective model that can make the distributed
representations of query related entities similar to the query in the vector
space. Hence, we propose three kinds of learning strategies, and the difference
between them mainly lies in how to deal with the relationship between an entity
and its description. We analyze the strengths and weaknesses of each learning
strategy and validate our methods on public datasets which contain four kinds
of named entities, i.e., movies, TV shows, restaurants and celebrities. The
experimental results indicate that our proposed methods can adapt to different
types of entity search queries, and outperform the current state-of-the-art
methods based on keyword matching and vanilla word2vec models. Besides, the
proposed methods can be trained fast and be easily extended to other similar
tasks.
| no_new_dataset | 0.942718 |
1611.00463 | Zahra Khatami | Zahra Khatami, Sungpack Hong, Jinsoo Lee, Siegfried Depner, Hassan
Chafi, J. Ramanujam, and Hartmut Kaiser | A Load-Balanced Parallel and Distributed Sorting Algorithm Implemented
with PGX.D | 8 pages, 12 figures | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sorting has been one of the most challenging studied problems in different
scientific researches. Although many techniques and algorithms have been
proposed on the theory of having efficient parallel sorting implementation,
however achieving desired performance on different types of the architectures
with large number of processors is still a challenging issue. Maximizing
parallelism level in applications can be achieved by minimizing overheads due
to load imbalance and waiting time due to memory latencies. In this paper, we
present a distributed sorting algorithm implemented in PGX.D, a fast
distributed graph processing system, which outperforms the Spark's distributed
sorting implementation by around 2x-3x by hiding communication latencies and
minimizing unnecessary overheads. Furthermore, it shows that the proposed PGX.D
sorting method handles dataset containing many duplicated data entries
efficiently and always results in keeping balanced workloads for different
input data distribution types.
| [
{
"version": "v1",
"created": "Wed, 2 Nov 2016 03:56:31 GMT"
},
{
"version": "v2",
"created": "Sat, 14 Jan 2017 20:17:32 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Khatami",
"Zahra",
""
],
[
"Hong",
"Sungpack",
""
],
[
"Lee",
"Jinsoo",
""
],
[
"Depner",
"Siegfried",
""
],
[
"Chafi",
"Hassan",
""
],
[
"Ramanujam",
"J.",
""
],
[
"Kaiser",
"Hartmut",
""
]
] | TITLE: A Load-Balanced Parallel and Distributed Sorting Algorithm Implemented
with PGX.D
ABSTRACT: Sorting has been one of the most challenging studied problems in different
scientific researches. Although many techniques and algorithms have been
proposed on the theory of having efficient parallel sorting implementation,
however achieving desired performance on different types of the architectures
with large number of processors is still a challenging issue. Maximizing
parallelism level in applications can be achieved by minimizing overheads due
to load imbalance and waiting time due to memory latencies. In this paper, we
present a distributed sorting algorithm implemented in PGX.D, a fast
distributed graph processing system, which outperforms the Spark's distributed
sorting implementation by around 2x-3x by hiding communication latencies and
minimizing unnecessary overheads. Furthermore, it shows that the proposed PGX.D
sorting method handles dataset containing many duplicated data entries
efficiently and always results in keeping balanced workloads for different
input data distribution types.
| no_new_dataset | 0.940188 |
1611.08069 | Bo Li | Bo Li | 3D Fully Convolutional Network for Vehicle Detection in Point Cloud | null | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 2D fully convolutional network has been recently successfully applied to
object detection from images. In this paper, we extend the fully convolutional
network based detection techniques to 3D and apply it to point cloud data. The
proposed approach is verified on the task of vehicle detection from lidar point
cloud for autonomous driving. Experiments on the KITTI dataset shows a
significant performance improvement over the previous point cloud based
detection approaches.
| [
{
"version": "v1",
"created": "Thu, 24 Nov 2016 05:06:05 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Jan 2017 05:56:01 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Li",
"Bo",
""
]
] | TITLE: 3D Fully Convolutional Network for Vehicle Detection in Point Cloud
ABSTRACT: 2D fully convolutional network has been recently successfully applied to
object detection from images. In this paper, we extend the fully convolutional
network based detection techniques to 3D and apply it to point cloud data. The
proposed approach is verified on the task of vehicle detection from lidar point
cloud for autonomous driving. Experiments on the KITTI dataset shows a
significant performance improvement over the previous point cloud based
detection approaches.
| no_new_dataset | 0.952397 |
1701.01500 | Haiqiang Wang | Haiqiang Wang, Ioannis Katsavounidis, Jiantong Zhou, Jeonghoon Park,
Shawmin Lei, Xin Zhou, Man-On Pun, Xin Jin, Ronggang Wang, Xu Wang, Yun
Zhang, Jiwu Huang, Sam Kwong and C.-C. Jay Kuo | VideoSet: A Large-Scale Compressed Video Quality Dataset Based on JND
Measurement | null | null | null | null | cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new methodology to measure coded image/video quality using the
just-noticeable-difference (JND) idea was proposed. Several small JND-based
image/video quality datasets were released by the Media Communications Lab at
the University of Southern California. In this work, we present an effort to
build a large-scale JND-based coded video quality dataset. The dataset consists
of 220 5-second sequences in four resolutions (i.e., $1920 \times 1080$, $1280
\times 720$, $960 \times 540$ and $640 \times 360$). For each of the 880 video
clips, we encode it using the H.264 codec with $QP=1, \cdots, 51$ and measure
the first three JND points with 30+ subjects. The dataset is called the
"VideoSet", which is an acronym for "Video Subject Evaluation Test (SET)". This
work describes the subjective test procedure, detection and removal of outlying
measured data, and the properties of collected JND data. Finally, the
significance and implications of the VideoSet to future video coding research
and standardization efforts are pointed out. All source/coded video clips as
well as measured JND data included in the VideoSet are available to the public
in the IEEE DataPort.
| [
{
"version": "v1",
"created": "Thu, 5 Jan 2017 23:14:01 GMT"
},
{
"version": "v2",
"created": "Sun, 15 Jan 2017 04:30:59 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Wang",
"Haiqiang",
""
],
[
"Katsavounidis",
"Ioannis",
""
],
[
"Zhou",
"Jiantong",
""
],
[
"Park",
"Jeonghoon",
""
],
[
"Lei",
"Shawmin",
""
],
[
"Zhou",
"Xin",
""
],
[
"Pun",
"Man-On",
""
],
[
"Jin",
"Xin",
""
],
[
"Wang",
"Ronggang",
""
],
[
"Wang",
"Xu",
""
],
[
"Zhang",
"Yun",
""
],
[
"Huang",
"Jiwu",
""
],
[
"Kwong",
"Sam",
""
],
[
"Kuo",
"C. -C. Jay",
""
]
] | TITLE: VideoSet: A Large-Scale Compressed Video Quality Dataset Based on JND
Measurement
ABSTRACT: A new methodology to measure coded image/video quality using the
just-noticeable-difference (JND) idea was proposed. Several small JND-based
image/video quality datasets were released by the Media Communications Lab at
the University of Southern California. In this work, we present an effort to
build a large-scale JND-based coded video quality dataset. The dataset consists
of 220 5-second sequences in four resolutions (i.e., $1920 \times 1080$, $1280
\times 720$, $960 \times 540$ and $640 \times 360$). For each of the 880 video
clips, we encode it using the H.264 codec with $QP=1, \cdots, 51$ and measure
the first three JND points with 30+ subjects. The dataset is called the
"VideoSet", which is an acronym for "Video Subject Evaluation Test (SET)". This
work describes the subjective test procedure, detection and removal of outlying
measured data, and the properties of collected JND data. Finally, the
significance and implications of the VideoSet to future video coding research
and standardization efforts are pointed out. All source/coded video clips as
well as measured JND data included in the VideoSet are available to the public
in the IEEE DataPort.
| new_dataset | 0.96395 |
1701.03162 | Yifan Yang | Yifan Yang and Tian Qin and Yu-Heng Lei | Real-time eSports Match Result Prediction | 8 pages, 8 figures | null | null | null | stat.AP cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we try to predict the winning team of a match in the
multiplayer eSports game Dota 2. To address the weaknesses of previous work, we
consider more aspects of prior (pre-match) features from individual players'
match history, as well as real-time (during-match) features at each minute as
the match progresses. We use logistic regression, the proposed Attribute
Sequence Model, and their combinations as the prediction models. In a dataset
of 78362 matches where 20631 matches contain replay data, our experiments show
that adding more aspects of prior features improves accuracy from 58.69% to
71.49%, and introducing real-time features achieves up to 93.73% accuracy when
predicting at the 40th minute.
| [
{
"version": "v1",
"created": "Sat, 10 Dec 2016 06:30:25 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Yang",
"Yifan",
""
],
[
"Qin",
"Tian",
""
],
[
"Lei",
"Yu-Heng",
""
]
] | TITLE: Real-time eSports Match Result Prediction
ABSTRACT: In this paper, we try to predict the winning team of a match in the
multiplayer eSports game Dota 2. To address the weaknesses of previous work, we
consider more aspects of prior (pre-match) features from individual players'
match history, as well as real-time (during-match) features at each minute as
the match progresses. We use logistic regression, the proposed Attribute
Sequence Model, and their combinations as the prediction models. In a dataset
of 78362 matches where 20631 matches contain replay data, our experiments show
that adding more aspects of prior features improves accuracy from 58.69% to
71.49%, and introducing real-time features achieves up to 93.73% accuracy when
predicting at the 40th minute.
| no_new_dataset | 0.807916 |
1701.03869 | Wenwen Ding | Wenwen Ding, Kai Liu | Learning Linear Dynamical Systems with High-Order Tensor Data for
Skeleton based Action Recognition | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, there has been renewed interest in developing methods for
skeleton-based human action recognition. A skeleton sequence can be naturally
represented as a high-order tensor time series. In this paper, we model and
analyze tensor time series with Linear Dynamical System (LDS) which is the most
common for encoding spatio-temporal time-series data in various disciplines dut
to its relative simplicity and efficiency. However, the traditional LDS treats
the latent and observation state at each frame of video as a column vector.
Such a vector representation fails to take into account the curse of
dimensionality as well as valuable structural information with human action.
Considering this fact, we propose generalized Linear Dynamical System (gLDS)
for modeling tensor observation in the time series and employ Tucker
decomposition to estimate the LDS parameters as action descriptors. Therefore,
an action can be represented as a subspace corresponding to a point on a
Grassmann manifold. Then we perform classification using dictionary learning
and sparse coding over Grassmann manifold. Experiments on MSR Action3D Dataset,
UCF Kinect Dataset and Northwestern-UCLA Multiview Action3D Dataset demonstrate
that our proposed method achieves superior performance to the state-of-the-art
algorithms.
| [
{
"version": "v1",
"created": "Sat, 14 Jan 2017 02:07:23 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Ding",
"Wenwen",
""
],
[
"Liu",
"Kai",
""
]
] | TITLE: Learning Linear Dynamical Systems with High-Order Tensor Data for
Skeleton based Action Recognition
ABSTRACT: In recent years, there has been renewed interest in developing methods for
skeleton-based human action recognition. A skeleton sequence can be naturally
represented as a high-order tensor time series. In this paper, we model and
analyze tensor time series with Linear Dynamical System (LDS) which is the most
common for encoding spatio-temporal time-series data in various disciplines dut
to its relative simplicity and efficiency. However, the traditional LDS treats
the latent and observation state at each frame of video as a column vector.
Such a vector representation fails to take into account the curse of
dimensionality as well as valuable structural information with human action.
Considering this fact, we propose generalized Linear Dynamical System (gLDS)
for modeling tensor observation in the time series and employ Tucker
decomposition to estimate the LDS parameters as action descriptors. Therefore,
an action can be represented as a subspace corresponding to a point on a
Grassmann manifold. Then we perform classification using dictionary learning
and sparse coding over Grassmann manifold. Experiments on MSR Action3D Dataset,
UCF Kinect Dataset and Northwestern-UCLA Multiview Action3D Dataset demonstrate
that our proposed method achieves superior performance to the state-of-the-art
algorithms.
| no_new_dataset | 0.950549 |
1701.03882 | Natalia Antropova Natalia Antropova | Natalia Antropova, Benjamin Huynh, Maryellen Giger | Multi-task Learning in the Computerized Diagnosis of Breast Cancer on
DCE-MRIs | null | null | null | null | physics.med-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hand-crafted features extracted from dynamic contrast-enhanced magnetic
resonance images (DCE-MRIs) have shown strong predictive abilities in
characterization of breast lesions. However, heterogeneity across medical image
datasets hinders the generalizability of these features. One of the sources of
the heterogeneity is the variation of MR scanner magnet strength, which has a
strong influence on image quality, leading to variations in the extracted image
features. Thus, statistical decision algorithms need to account for such data
heterogeneity. Despite the variations, we hypothesize that there exist
underlying relationships between the features extracted from the datasets
acquired with different magnet strength MR scanners. We compared the use of a
multi-task learning (MTL) method that incorporates those relationships during
the classifier training to support vector machines run on a merged dataset that
includes cases with various MRI strength images. As a result, higher predictive
power is achieved with the MTL method.
| [
{
"version": "v1",
"created": "Sat, 14 Jan 2017 05:55:02 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Antropova",
"Natalia",
""
],
[
"Huynh",
"Benjamin",
""
],
[
"Giger",
"Maryellen",
""
]
] | TITLE: Multi-task Learning in the Computerized Diagnosis of Breast Cancer on
DCE-MRIs
ABSTRACT: Hand-crafted features extracted from dynamic contrast-enhanced magnetic
resonance images (DCE-MRIs) have shown strong predictive abilities in
characterization of breast lesions. However, heterogeneity across medical image
datasets hinders the generalizability of these features. One of the sources of
the heterogeneity is the variation of MR scanner magnet strength, which has a
strong influence on image quality, leading to variations in the extracted image
features. Thus, statistical decision algorithms need to account for such data
heterogeneity. Despite the variations, we hypothesize that there exist
underlying relationships between the features extracted from the datasets
acquired with different magnet strength MR scanners. We compared the use of a
multi-task learning (MTL) method that incorporates those relationships during
the classifier training to support vector machines run on a merged dataset that
includes cases with various MRI strength images. As a result, higher predictive
power is achieved with the MTL method.
| no_new_dataset | 0.949201 |
1701.03937 | Tuan Tran | Tuan Tran, Tu Ngoc Nguyen | Hedera: Scalable Indexing and Exploring Entities in Wikipedia Revision
History | Pubished via CEUR-WS.org/Vol-1272 | null | null | null | cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Much of work in semantic web relying on Wikipedia as the main source of
knowledge often work on static snapshots of the dataset. The full history of
Wikipedia revisions, while contains much more useful information, is still
difficult to access due to its exceptional volume. To enable further research
on this collection, we developed a tool, named Hedera, that efficiently
extracts semantic information from Wikipedia revision history datasets. Hedera
exploits Map-Reduce paradigm to achieve rapid extraction, it is able to handle
one entire Wikipedia articles revision history within a day in a medium-scale
cluster, and supports flexible data structures for various kinds of semantic
web study.
| [
{
"version": "v1",
"created": "Sat, 14 Jan 2017 15:47:06 GMT"
}
] | 2017-01-17T00:00:00 | [
[
"Tran",
"Tuan",
""
],
[
"Nguyen",
"Tu Ngoc",
""
]
] | TITLE: Hedera: Scalable Indexing and Exploring Entities in Wikipedia Revision
History
ABSTRACT: Much of work in semantic web relying on Wikipedia as the main source of
knowledge often work on static snapshots of the dataset. The full history of
Wikipedia revisions, while contains much more useful information, is still
difficult to access due to its exceptional volume. To enable further research
on this collection, we developed a tool, named Hedera, that efficiently
extracts semantic information from Wikipedia revision history datasets. Hedera
exploits Map-Reduce paradigm to achieve rapid extraction, it is able to handle
one entire Wikipedia articles revision history within a day in a medium-scale
cluster, and supports flexible data structures for various kinds of semantic
web study.
| no_new_dataset | 0.943504 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.