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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1512.01525 | Yezhou Yang | Yezhou Yang and Yiannis Aloimonos and Cornelia Fermuller and Eren
Erdal Aksoy | Learning the Semantics of Manipulation Action | null | The 53rd Annual Meeting of the Association for Computational
Linguistics (ACL) 1 (2015) 676-686 | null | null | cs.RO cs.CL cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we present a formal computational framework for modeling
manipulation actions. The introduced formalism leads to semantics of
manipulation action and has applications to both observing and understanding
human manipulation actions as well as executing them with a robotic mechanism
(e.g. a humanoid robot). It is based on a Combinatory Categorial Grammar. The
goal of the introduced framework is to: (1) represent manipulation actions with
both syntax and semantic parts, where the semantic part employs
$\lambda$-calculus; (2) enable a probabilistic semantic parsing schema to learn
the $\lambda$-calculus representation of manipulation action from an annotated
action corpus of videos; (3) use (1) and (2) to develop a system that visually
observes manipulation actions and understands their meaning while it can reason
beyond observations using propositional logic and axiom schemata. The
experiments conducted on a public available large manipulation action dataset
validate the theoretical framework and our implementation.
| [
{
"version": "v1",
"created": "Fri, 4 Dec 2015 20:00:08 GMT"
}
] | 2015-12-07T00:00:00 | [
[
"Yang",
"Yezhou",
""
],
[
"Aloimonos",
"Yiannis",
""
],
[
"Fermuller",
"Cornelia",
""
],
[
"Aksoy",
"Eren Erdal",
""
]
] | TITLE: Learning the Semantics of Manipulation Action
ABSTRACT: In this paper we present a formal computational framework for modeling
manipulation actions. The introduced formalism leads to semantics of
manipulation action and has applications to both observing and understanding
human manipulation actions as well as executing them with a robotic mechanism
(e.g. a humanoid robot). It is based on a Combinatory Categorial Grammar. The
goal of the introduced framework is to: (1) represent manipulation actions with
both syntax and semantic parts, where the semantic part employs
$\lambda$-calculus; (2) enable a probabilistic semantic parsing schema to learn
the $\lambda$-calculus representation of manipulation action from an annotated
action corpus of videos; (3) use (1) and (2) to develop a system that visually
observes manipulation actions and understands their meaning while it can reason
beyond observations using propositional logic and axiom schemata. The
experiments conducted on a public available large manipulation action dataset
validate the theoretical framework and our implementation.
| no_new_dataset | 0.942507 |
1510.06925 | Leigh Robinson | Leigh Robinson, Benjamin Graham | Confusing Deep Convolution Networks by Relabelling | Submitted to BMVC 2015 | null | null | null | cs.CV cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep convolutional neural networks have become the gold standard for image
recognition tasks, demonstrating many current state-of-the-art results and even
achieving near-human level performance on some tasks. Despite this fact it has
been shown that their strong generalisation qualities can be fooled to
misclassify previously correctly classified natural images and give erroneous
high confidence classifications to nonsense synthetic images. In this paper we
extend that work, by presenting a straightforward way to perturb an image in
such a way as to cause it to acquire any other label from within the dataset
while leaving this perturbed image visually indistinguishable from the
original.
| [
{
"version": "v1",
"created": "Fri, 23 Oct 2015 13:02:55 GMT"
},
{
"version": "v2",
"created": "Thu, 3 Dec 2015 18:38:08 GMT"
}
] | 2015-12-04T00:00:00 | [
[
"Robinson",
"Leigh",
""
],
[
"Graham",
"Benjamin",
""
]
] | TITLE: Confusing Deep Convolution Networks by Relabelling
ABSTRACT: Deep convolutional neural networks have become the gold standard for image
recognition tasks, demonstrating many current state-of-the-art results and even
achieving near-human level performance on some tasks. Despite this fact it has
been shown that their strong generalisation qualities can be fooled to
misclassify previously correctly classified natural images and give erroneous
high confidence classifications to nonsense synthetic images. In this paper we
extend that work, by presenting a straightforward way to perturb an image in
such a way as to cause it to acquire any other label from within the dataset
while leaving this perturbed image visually indistinguishable from the
original.
| no_new_dataset | 0.947235 |
1511.06440 | Ilya Sutskever | Ilya Sutskever, Rafal Jozefowicz, Karol Gregor, Danilo Rezende, Tim
Lillicrap, Oriol Vinyals | Towards Principled Unsupervised Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | General unsupervised learning is a long-standing conceptual problem in
machine learning. Supervised learning is successful because it can be solved by
the minimization of the training error cost function. Unsupervised learning is
not as successful, because the unsupervised objective may be unrelated to the
supervised task of interest. For an example, density modelling and
reconstruction have often been used for unsupervised learning, but they did not
produced the sought-after performance gains, because they have no knowledge of
the supervised tasks.
In this paper, we present an unsupervised cost function which we name the
Output Distribution Matching (ODM) cost, which measures a divergence between
the distribution of predictions and distributions of labels. The ODM cost is
appealing because it is consistent with the supervised cost in the following
sense: a perfect supervised classifier is also perfect according to the ODM
cost. Therefore, by aggressively optimizing the ODM cost, we are almost
guaranteed to improve our supervised performance whenever the space of possible
predictions is exponentially large.
We demonstrate that the ODM cost works well on number of small and
semi-artificial datasets using no (or almost no) labelled training cases.
Finally, we show that the ODM cost can be used for one-shot domain adaptation,
which allows the model to classify inputs that differ from the input
distribution in significant ways without the need for prior exposure to the new
domain.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 23:04:23 GMT"
},
{
"version": "v2",
"created": "Thu, 3 Dec 2015 17:24:22 GMT"
}
] | 2015-12-04T00:00:00 | [
[
"Sutskever",
"Ilya",
""
],
[
"Jozefowicz",
"Rafal",
""
],
[
"Gregor",
"Karol",
""
],
[
"Rezende",
"Danilo",
""
],
[
"Lillicrap",
"Tim",
""
],
[
"Vinyals",
"Oriol",
""
]
] | TITLE: Towards Principled Unsupervised Learning
ABSTRACT: General unsupervised learning is a long-standing conceptual problem in
machine learning. Supervised learning is successful because it can be solved by
the minimization of the training error cost function. Unsupervised learning is
not as successful, because the unsupervised objective may be unrelated to the
supervised task of interest. For an example, density modelling and
reconstruction have often been used for unsupervised learning, but they did not
produced the sought-after performance gains, because they have no knowledge of
the supervised tasks.
In this paper, we present an unsupervised cost function which we name the
Output Distribution Matching (ODM) cost, which measures a divergence between
the distribution of predictions and distributions of labels. The ODM cost is
appealing because it is consistent with the supervised cost in the following
sense: a perfect supervised classifier is also perfect according to the ODM
cost. Therefore, by aggressively optimizing the ODM cost, we are almost
guaranteed to improve our supervised performance whenever the space of possible
predictions is exponentially large.
We demonstrate that the ODM cost works well on number of small and
semi-artificial datasets using no (or almost no) labelled training cases.
Finally, we show that the ODM cost can be used for one-shot domain adaptation,
which allows the model to classify inputs that differ from the input
distribution in significant ways without the need for prior exposure to the new
domain.
| no_new_dataset | 0.943556 |
1512.00901 | Mingrui Yang | Mingrui Yang, Frank de Hoog, Yuqi Fan, and Wen Hu | Compressive hyperspectral imaging via adaptive sampling and dictionary
learning | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a new sampling strategy for hyperspectral signals
that is based on dictionary learning and singular value decomposition (SVD).
Specifically, we first learn a sparsifying dictionary from training spectral
data using dictionary learning. We then perform an SVD on the dictionary and
use the first few left singular vectors as the rows of the measurement matrix
to obtain the compressive measurements for reconstruction. The proposed method
provides significant improvement over the conventional compressive sensing
approaches. The reconstruction performance is further improved by
reconditioning the sensing matrix using matrix balancing. We also demonstrate
that the combination of dictionary learning and SVD is robust by applying them
to different datasets.
| [
{
"version": "v1",
"created": "Wed, 2 Dec 2015 23:13:04 GMT"
}
] | 2015-12-04T00:00:00 | [
[
"Yang",
"Mingrui",
""
],
[
"de Hoog",
"Frank",
""
],
[
"Fan",
"Yuqi",
""
],
[
"Hu",
"Wen",
""
]
] | TITLE: Compressive hyperspectral imaging via adaptive sampling and dictionary
learning
ABSTRACT: In this paper, we propose a new sampling strategy for hyperspectral signals
that is based on dictionary learning and singular value decomposition (SVD).
Specifically, we first learn a sparsifying dictionary from training spectral
data using dictionary learning. We then perform an SVD on the dictionary and
use the first few left singular vectors as the rows of the measurement matrix
to obtain the compressive measurements for reconstruction. The proposed method
provides significant improvement over the conventional compressive sensing
approaches. The reconstruction performance is further improved by
reconditioning the sensing matrix using matrix balancing. We also demonstrate
that the combination of dictionary learning and SVD is robust by applying them
to different datasets.
| no_new_dataset | 0.948394 |
1512.01055 | Ibrahim Radwan Dr. | Ibrahim Radwan, Abhinav Dhall, Roland Goecke | Occlusion-Aware Human Pose Estimation with Mixtures of Sub-Trees | 12 pages, 5 figures and 3 Tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study the problem of learning a model for human pose
estimation as mixtures of compositional sub-trees in two layers of prediction.
This involves estimating the pose of a sub-tree followed by identifying the
relationships between sub-trees that are used to handle occlusions between
different parts. The mixtures of the sub-trees are learnt utilising both
geometric and appearance distances. The Chow-Liu (CL) algorithm is recursively
applied to determine the inter-relations between the nodes and to build the
structure of the sub-trees. These structures are used to learn the latent
parameters of the sub-trees and the inference is done using a standard belief
propagation technique. The proposed method handles occlusions during the
inference process by identifying overlapping regions between different
sub-trees and introducing a penalty term for overlapping parts. Experiments are
performed on three different datasets: the Leeds Sports, Image Parse and UIUC
People datasets. The results show the robustness of the proposed method to
occlusions over the state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Thu, 3 Dec 2015 12:25:33 GMT"
}
] | 2015-12-04T00:00:00 | [
[
"Radwan",
"Ibrahim",
""
],
[
"Dhall",
"Abhinav",
""
],
[
"Goecke",
"Roland",
""
]
] | TITLE: Occlusion-Aware Human Pose Estimation with Mixtures of Sub-Trees
ABSTRACT: In this paper, we study the problem of learning a model for human pose
estimation as mixtures of compositional sub-trees in two layers of prediction.
This involves estimating the pose of a sub-tree followed by identifying the
relationships between sub-trees that are used to handle occlusions between
different parts. The mixtures of the sub-trees are learnt utilising both
geometric and appearance distances. The Chow-Liu (CL) algorithm is recursively
applied to determine the inter-relations between the nodes and to build the
structure of the sub-trees. These structures are used to learn the latent
parameters of the sub-trees and the inference is done using a standard belief
propagation technique. The proposed method handles occlusions during the
inference process by identifying overlapping regions between different
sub-trees and introducing a penalty term for overlapping parts. Experiments are
performed on three different datasets: the Leeds Sports, Image Parse and UIUC
People datasets. The results show the robustness of the proposed method to
occlusions over the state-of-the-art approaches.
| no_new_dataset | 0.943452 |
1512.00537 | Anja Gruenheid | Anja Gruenheid and Besmira Nushi and Tim Kraska and Wolfgang
Gatterbauer and Donald Kossmann | Fault-Tolerant Entity Resolution with the Crowd | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, crowdsourcing is increasingly applied as a means to enhance
data quality. Although the crowd generates insightful information especially
for complex problems such as entity resolution (ER), the output quality of
crowd workers is often noisy. That is, workers may unintentionally generate
false or contradicting data even for simple tasks. The challenge that we
address in this paper is how to minimize the cost for task requesters while
maximizing ER result quality under the assumption of unreliable input from the
crowd. For that purpose, we first establish how to deduce a consistent ER
solution from noisy worker answers as part of the data interpretation problem.
We then focus on the next-crowdsource problem which is to find the next task
that maximizes the information gain of the ER result for the minimal additional
cost. We compare our robust data interpretation strategies to alternative
state-of-the-art approaches that do not incorporate the notion of
fault-tolerance, i.e., the robustness to noise. In our experimental evaluation
we show that our approaches yield a quality improvement of at least 20% for two
real-world datasets. Furthermore, we examine task-to-worker assignment
strategies as well as task parallelization techniques in terms of their cost
and quality trade-offs in this paper. Based on both synthetic and crowdsourced
datasets, we then draw conclusions on how to minimize cost while maintaining
high quality ER results.
| [
{
"version": "v1",
"created": "Wed, 2 Dec 2015 01:03:47 GMT"
}
] | 2015-12-03T00:00:00 | [
[
"Gruenheid",
"Anja",
""
],
[
"Nushi",
"Besmira",
""
],
[
"Kraska",
"Tim",
""
],
[
"Gatterbauer",
"Wolfgang",
""
],
[
"Kossmann",
"Donald",
""
]
] | TITLE: Fault-Tolerant Entity Resolution with the Crowd
ABSTRACT: In recent years, crowdsourcing is increasingly applied as a means to enhance
data quality. Although the crowd generates insightful information especially
for complex problems such as entity resolution (ER), the output quality of
crowd workers is often noisy. That is, workers may unintentionally generate
false or contradicting data even for simple tasks. The challenge that we
address in this paper is how to minimize the cost for task requesters while
maximizing ER result quality under the assumption of unreliable input from the
crowd. For that purpose, we first establish how to deduce a consistent ER
solution from noisy worker answers as part of the data interpretation problem.
We then focus on the next-crowdsource problem which is to find the next task
that maximizes the information gain of the ER result for the minimal additional
cost. We compare our robust data interpretation strategies to alternative
state-of-the-art approaches that do not incorporate the notion of
fault-tolerance, i.e., the robustness to noise. In our experimental evaluation
we show that our approaches yield a quality improvement of at least 20% for two
real-world datasets. Furthermore, we examine task-to-worker assignment
strategies as well as task parallelization techniques in terms of their cost
and quality trade-offs in this paper. Based on both synthetic and crowdsourced
datasets, we then draw conclusions on how to minimize cost while maintaining
high quality ER results.
| no_new_dataset | 0.953665 |
1512.00596 | Ira Kemelmacher-Shlizerman | Ira Kemelmacher-Shlizerman and Steve Seitz and Daniel Miller and Evan
Brossard | The MegaFace Benchmark: 1 Million Faces for Recognition at Scale | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent face recognition experiments on a major benchmark LFW show stunning
performance--a number of algorithms achieve near to perfect score, surpassing
human recognition rates. In this paper, we advocate evaluations at the million
scale (LFW includes only 13K photos of 5K people). To this end, we have
assembled the MegaFace dataset and created the first MegaFace challenge. Our
dataset includes One Million photos that capture more than 690K different
individuals. The challenge evaluates performance of algorithms with increasing
numbers of distractors (going from 10 to 1M) in the gallery set. We present
both identification and verification performance, evaluate performance with
respect to pose and a person's age, and compare as a function of training data
size (number of photos and people). We report results of state of the art and
baseline algorithms. Our key observations are that testing at the million scale
reveals big performance differences (of algorithms that perform similarly well
on smaller scale) and that age invariant recognition as well as pose are still
challenging for most. The MegaFace dataset, baseline code, and evaluation
scripts, are all publicly released for further experimentations at:
megaface.cs.washington.edu.
| [
{
"version": "v1",
"created": "Wed, 2 Dec 2015 07:17:54 GMT"
}
] | 2015-12-03T00:00:00 | [
[
"Kemelmacher-Shlizerman",
"Ira",
""
],
[
"Seitz",
"Steve",
""
],
[
"Miller",
"Daniel",
""
],
[
"Brossard",
"Evan",
""
]
] | TITLE: The MegaFace Benchmark: 1 Million Faces for Recognition at Scale
ABSTRACT: Recent face recognition experiments on a major benchmark LFW show stunning
performance--a number of algorithms achieve near to perfect score, surpassing
human recognition rates. In this paper, we advocate evaluations at the million
scale (LFW includes only 13K photos of 5K people). To this end, we have
assembled the MegaFace dataset and created the first MegaFace challenge. Our
dataset includes One Million photos that capture more than 690K different
individuals. The challenge evaluates performance of algorithms with increasing
numbers of distractors (going from 10 to 1M) in the gallery set. We present
both identification and verification performance, evaluate performance with
respect to pose and a person's age, and compare as a function of training data
size (number of photos and people). We report results of state of the art and
baseline algorithms. Our key observations are that testing at the million scale
reveals big performance differences (of algorithms that perform similarly well
on smaller scale) and that age invariant recognition as well as pose are still
challenging for most. The MegaFace dataset, baseline code, and evaluation
scripts, are all publicly released for further experimentations at:
megaface.cs.washington.edu.
| new_dataset | 0.955486 |
1512.00682 | Davut Deniz Yavuz | Davut Deniz Yavuz, Osman Abul | Implicit Location Sharing Detection in Social Media from Short Turkish
Text | 6 pages | null | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Social media have become a significant venue for information sharing of live
updates. Users of social media are producing and sharing large amount of
personal data as a part of the live updates. A significant percentage of this
data contains location information that can be used by other people for many
purposes. Some of the social media users deliberately share their own location
information with other social network users. However, a large number of social
media users blindly or implicitly share their location without noticing it or
its possible consequences. Implicit location sharing is investigated in the
current paper. We perform a large scale study on implicit location sharing for
one of the most popular social media platform, namely Twitter. After a careful
study, we built a dataset of Turkish tweets and manually tagged them. Using
machine learning techniques we built classifiers that are able to classify
whether a given tweet contains implicit location sharing or not. The
classifiers are shown to be very accurate and efficient. Moreover, the best
classifier is employed as a browser add-on tool which warns the user whenever
an implicit location sharing is predicted from to be released tweet. The paper
provides the methodology and the technical analysis as well. Furthermore, it
discusses how these techniques can be extended to different social network
services and also to different languages.
| [
{
"version": "v1",
"created": "Wed, 2 Dec 2015 13:10:33 GMT"
}
] | 2015-12-03T00:00:00 | [
[
"Yavuz",
"Davut Deniz",
""
],
[
"Abul",
"Osman",
""
]
] | TITLE: Implicit Location Sharing Detection in Social Media from Short Turkish
Text
ABSTRACT: Social media have become a significant venue for information sharing of live
updates. Users of social media are producing and sharing large amount of
personal data as a part of the live updates. A significant percentage of this
data contains location information that can be used by other people for many
purposes. Some of the social media users deliberately share their own location
information with other social network users. However, a large number of social
media users blindly or implicitly share their location without noticing it or
its possible consequences. Implicit location sharing is investigated in the
current paper. We perform a large scale study on implicit location sharing for
one of the most popular social media platform, namely Twitter. After a careful
study, we built a dataset of Turkish tweets and manually tagged them. Using
machine learning techniques we built classifiers that are able to classify
whether a given tweet contains implicit location sharing or not. The
classifiers are shown to be very accurate and efficient. Moreover, the best
classifier is employed as a browser add-on tool which warns the user whenever
an implicit location sharing is predicted from to be released tweet. The paper
provides the methodology and the technical analysis as well. Furthermore, it
discusses how these techniques can be extended to different social network
services and also to different languages.
| new_dataset | 0.957437 |
1512.00728 | David Bamman | Philip Massey, Patrick Xia, David Bamman and Noah A. Smith | Annotating Character Relationships in Literary Texts | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present a dataset of manually annotated relationships between characters
in literary texts, in order to support the training and evaluation of automatic
methods for relation type prediction in this domain (Makazhanov et al., 2014;
Kokkinakis, 2013) and the broader computational analysis of literary character
(Elson et al., 2010; Bamman et al., 2014; Vala et al., 2015; Flekova and
Gurevych, 2015). In this work, we solicit annotations from workers on Amazon
Mechanical Turk for 109 texts ranging from Homer's _Iliad_ to Joyce's _Ulysses_
on four dimensions of interest: for a given pair of characters, we collect
judgments as to the coarse-grained category (professional, social, familial),
fine-grained category (friend, lover, parent, rival, employer), and affinity
(positive, negative, neutral) that describes their primary relationship in a
text. We do not assume that this relationship is static; we also collect
judgments as to whether it changes at any point in the course of the text.
| [
{
"version": "v1",
"created": "Wed, 2 Dec 2015 15:09:31 GMT"
}
] | 2015-12-03T00:00:00 | [
[
"Massey",
"Philip",
""
],
[
"Xia",
"Patrick",
""
],
[
"Bamman",
"David",
""
],
[
"Smith",
"Noah A.",
""
]
] | TITLE: Annotating Character Relationships in Literary Texts
ABSTRACT: We present a dataset of manually annotated relationships between characters
in literary texts, in order to support the training and evaluation of automatic
methods for relation type prediction in this domain (Makazhanov et al., 2014;
Kokkinakis, 2013) and the broader computational analysis of literary character
(Elson et al., 2010; Bamman et al., 2014; Vala et al., 2015; Flekova and
Gurevych, 2015). In this work, we solicit annotations from workers on Amazon
Mechanical Turk for 109 texts ranging from Homer's _Iliad_ to Joyce's _Ulysses_
on four dimensions of interest: for a given pair of characters, we collect
judgments as to the coarse-grained category (professional, social, familial),
fine-grained category (friend, lover, parent, rival, employer), and affinity
(positive, negative, neutral) that describes their primary relationship in a
text. We do not assume that this relationship is static; we also collect
judgments as to whether it changes at any point in the course of the text.
| new_dataset | 0.959875 |
1512.00747 | Agata Mosinska | Agata Mosinska, Raphael Sznitman, Przemys{\l}aw G{\l}owacki, Pascal
Fua | Active Learning for Delineation of Curvilinear Structures | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many recent delineation techniques owe much of their increased effectiveness
to path classification algorithms that make it possible to distinguish
promising paths from others. The downside of this development is that they
require annotated training data, which is tedious to produce.
In this paper, we propose an Active Learning approach that considerably
speeds up the annotation process. Unlike standard ones, it takes advantage of
the specificities of the delineation problem. It operates on a graph and can
reduce the training set size by up to 80% without compromising the
reconstruction quality.
We will show that our approach outperforms conventional ones on various
biomedical and natural image datasets, thus showing that it is broadly
applicable.
| [
{
"version": "v1",
"created": "Wed, 2 Dec 2015 15:57:59 GMT"
}
] | 2015-12-03T00:00:00 | [
[
"Mosinska",
"Agata",
""
],
[
"Sznitman",
"Raphael",
""
],
[
"Głowacki",
"Przemysław",
""
],
[
"Fua",
"Pascal",
""
]
] | TITLE: Active Learning for Delineation of Curvilinear Structures
ABSTRACT: Many recent delineation techniques owe much of their increased effectiveness
to path classification algorithms that make it possible to distinguish
promising paths from others. The downside of this development is that they
require annotated training data, which is tedious to produce.
In this paper, we propose an Active Learning approach that considerably
speeds up the annotation process. Unlike standard ones, it takes advantage of
the specificities of the delineation problem. It operates on a graph and can
reduce the training set size by up to 80% without compromising the
reconstruction quality.
We will show that our approach outperforms conventional ones on various
biomedical and natural image datasets, thus showing that it is broadly
applicable.
| no_new_dataset | 0.947381 |
1504.04660 | Neal Hurlburt | Neal Hurlburt and Steve Jaffey | A spectral optical flow method for determining velocities from digital
imagery | 12 pages, 5 figures. Submitted to Earth Science Informatics | Earth Science Informatics: Volume 8, Issue 4 (2015), Page 959-965 | 10.1007/s12145-015-0224-4 | null | cs.CV astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a method for determining surface flows from solar images based
upon optical flow techniques. We apply the method to sets of images obtained by
a variety of solar imagers to assess its performance. The {\tt opflow3d}
procedure is shown to extract accurate velocity estimates when provided perfect
test data and quickly generates results consistent with completely distinct
methods when applied on global scales. We also validate it in detail by
comparing it to an established method when applied to high-resolution datasets
and find that it provides comparable results without the need to tune, filter
or otherwise preprocess the images before its application.
| [
{
"version": "v1",
"created": "Fri, 17 Apr 2015 23:44:20 GMT"
}
] | 2015-12-02T00:00:00 | [
[
"Hurlburt",
"Neal",
""
],
[
"Jaffey",
"Steve",
""
]
] | TITLE: A spectral optical flow method for determining velocities from digital
imagery
ABSTRACT: We present a method for determining surface flows from solar images based
upon optical flow techniques. We apply the method to sets of images obtained by
a variety of solar imagers to assess its performance. The {\tt opflow3d}
procedure is shown to extract accurate velocity estimates when provided perfect
test data and quickly generates results consistent with completely distinct
methods when applied on global scales. We also validate it in detail by
comparing it to an established method when applied to high-resolution datasets
and find that it provides comparable results without the need to tune, filter
or otherwise preprocess the images before its application.
| no_new_dataset | 0.95253 |
1512.00001 | Stan Hatko | Stan Hatko | k-Nearest Neighbour Classification of Datasets with a Family of
Distances | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The $k$-nearest neighbour ($k$-NN) classifier is one of the oldest and most
important supervised learning algorithms for classifying datasets.
Traditionally the Euclidean norm is used as the distance for the $k$-NN
classifier. In this thesis we investigate the use of alternative distances for
the $k$-NN classifier.
We start by introducing some background notions in statistical machine
learning. We define the $k$-NN classifier and discuss Stone's theorem and the
proof that $k$-NN is universally consistent on the normed space $R^d$. We then
prove that $k$-NN is universally consistent if we take a sequence of random
norms (that are independent of the sample and the query) from a family of norms
that satisfies a particular boundedness condition. We extend this result by
replacing norms with distances based on uniformly locally Lipschitz functions
that satisfy certain conditions. We discuss the limitations of Stone's lemma
and Stone's theorem, particularly with respect to quasinorms and adaptively
choosing a distance for $k$-NN based on the labelled sample. We show the
universal consistency of a two stage $k$-NN type classifier where we select the
distance adaptively based on a split labelled sample and the query. We conclude
by giving some examples of improvements of the accuracy of classifying various
datasets using the above techniques.
| [
{
"version": "v1",
"created": "Sun, 29 Nov 2015 01:52:34 GMT"
}
] | 2015-12-02T00:00:00 | [
[
"Hatko",
"Stan",
""
]
] | TITLE: k-Nearest Neighbour Classification of Datasets with a Family of
Distances
ABSTRACT: The $k$-nearest neighbour ($k$-NN) classifier is one of the oldest and most
important supervised learning algorithms for classifying datasets.
Traditionally the Euclidean norm is used as the distance for the $k$-NN
classifier. In this thesis we investigate the use of alternative distances for
the $k$-NN classifier.
We start by introducing some background notions in statistical machine
learning. We define the $k$-NN classifier and discuss Stone's theorem and the
proof that $k$-NN is universally consistent on the normed space $R^d$. We then
prove that $k$-NN is universally consistent if we take a sequence of random
norms (that are independent of the sample and the query) from a family of norms
that satisfies a particular boundedness condition. We extend this result by
replacing norms with distances based on uniformly locally Lipschitz functions
that satisfy certain conditions. We discuss the limitations of Stone's lemma
and Stone's theorem, particularly with respect to quasinorms and adaptively
choosing a distance for $k$-NN based on the labelled sample. We show the
universal consistency of a two stage $k$-NN type classifier where we select the
distance adaptively based on a split labelled sample and the query. We conclude
by giving some examples of improvements of the accuracy of classifying various
datasets using the above techniques.
| no_new_dataset | 0.946892 |
1512.00066 | Edgar Solomonik | Edgar Solomonik and Torsten Hoefler | Sparse Tensor Algebra as a Parallel Programming Model | null | null | null | null | cs.MS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dense and sparse tensors allow the representation of most bulk data
structures in computational science applications. We show that sparse tensor
algebra can also be used to express many of the transformations on these
datasets, especially those which are parallelizable. Tensor computations are a
natural generalization of matrix and graph computations. We extend the usual
basic operations of tensor summation and contraction to arbitrary functions,
and further operations such as reductions and mapping. The expression of these
transformations in a high-level sparse linear algebra domain specific language
allows our framework to understand their properties at runtime to select the
preferred communication-avoiding algorithm. To demonstrate the efficacy of our
approach, we show how key graph algorithms as well as common numerical kernels
can be succinctly expressed using our interface and provide performance results
of a general library implementation.
| [
{
"version": "v1",
"created": "Mon, 30 Nov 2015 22:08:23 GMT"
}
] | 2015-12-02T00:00:00 | [
[
"Solomonik",
"Edgar",
""
],
[
"Hoefler",
"Torsten",
""
]
] | TITLE: Sparse Tensor Algebra as a Parallel Programming Model
ABSTRACT: Dense and sparse tensors allow the representation of most bulk data
structures in computational science applications. We show that sparse tensor
algebra can also be used to express many of the transformations on these
datasets, especially those which are parallelizable. Tensor computations are a
natural generalization of matrix and graph computations. We extend the usual
basic operations of tensor summation and contraction to arbitrary functions,
and further operations such as reductions and mapping. The expression of these
transformations in a high-level sparse linear algebra domain specific language
allows our framework to understand their properties at runtime to select the
preferred communication-avoiding algorithm. To demonstrate the efficacy of our
approach, we show how key graph algorithms as well as common numerical kernels
can be succinctly expressed using our interface and provide performance results
of a general library implementation.
| no_new_dataset | 0.939969 |
1512.00112 | Shashank Srivastava | Shashank Srivastava, Snigdha Chaturvedi and Tom Mitchell | Inferring Interpersonal Relations in Narrative Summaries | null | null | null | null | cs.CL cs.AI cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Characterizing relationships between people is fundamental for the
understanding of narratives. In this work, we address the problem of inferring
the polarity of relationships between people in narrative summaries. We
formulate the problem as a joint structured prediction for each narrative, and
present a model that combines evidence from linguistic and semantic features,
as well as features based on the structure of the social community in the text.
We also provide a clustering-based approach that can exploit regularities in
narrative types. e.g., learn an affinity for love-triangles in romantic
stories. On a dataset of movie summaries from Wikipedia, our structured models
provide more than a 30% error-reduction over a competitive baseline that
considers pairs of characters in isolation.
| [
{
"version": "v1",
"created": "Tue, 1 Dec 2015 01:11:46 GMT"
}
] | 2015-12-02T00:00:00 | [
[
"Srivastava",
"Shashank",
""
],
[
"Chaturvedi",
"Snigdha",
""
],
[
"Mitchell",
"Tom",
""
]
] | TITLE: Inferring Interpersonal Relations in Narrative Summaries
ABSTRACT: Characterizing relationships between people is fundamental for the
understanding of narratives. In this work, we address the problem of inferring
the polarity of relationships between people in narrative summaries. We
formulate the problem as a joint structured prediction for each narrative, and
present a model that combines evidence from linguistic and semantic features,
as well as features based on the structure of the social community in the text.
We also provide a clustering-based approach that can exploit regularities in
narrative types. e.g., learn an affinity for love-triangles in romantic
stories. On a dataset of movie summaries from Wikipedia, our structured models
provide more than a 30% error-reduction over a competitive baseline that
considers pairs of characters in isolation.
| no_new_dataset | 0.95018 |
1512.00130 | Tyng-Luh Liu | Tsung-Yu Lin, Tsung-Wei Ke, Tyng-Luh Liu | Implicit Sparse Code Hashing | 9 pages, 1 figure | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We address the problem of converting large-scale high-dimensional image data
into binary codes so that approximate nearest-neighbor search over them can be
efficiently performed. Different from most of the existing unsupervised
approaches for yielding binary codes, our method is based on a
dimensionality-reduction criterion that its resulting mapping is designed to
preserve the image relationships entailed by the inner products of sparse
codes, rather than those implied by the Euclidean distances in the ambient
space. While the proposed formulation does not require computing any sparse
codes, the underlying computation model still inevitably involves solving an
unmanageable eigenproblem when extremely high-dimensional descriptors are used.
To overcome the difficulty, we consider the column-sampling technique and
presume a special form of rotation matrix to facilitate subproblem
decomposition. We test our method on several challenging image datasets and
demonstrate its effectiveness by comparing with state-of-the-art binary coding
techniques.
| [
{
"version": "v1",
"created": "Tue, 1 Dec 2015 03:12:09 GMT"
}
] | 2015-12-02T00:00:00 | [
[
"Lin",
"Tsung-Yu",
""
],
[
"Ke",
"Tsung-Wei",
""
],
[
"Liu",
"Tyng-Luh",
""
]
] | TITLE: Implicit Sparse Code Hashing
ABSTRACT: We address the problem of converting large-scale high-dimensional image data
into binary codes so that approximate nearest-neighbor search over them can be
efficiently performed. Different from most of the existing unsupervised
approaches for yielding binary codes, our method is based on a
dimensionality-reduction criterion that its resulting mapping is designed to
preserve the image relationships entailed by the inner products of sparse
codes, rather than those implied by the Euclidean distances in the ambient
space. While the proposed formulation does not require computing any sparse
codes, the underlying computation model still inevitably involves solving an
unmanageable eigenproblem when extremely high-dimensional descriptors are used.
To overcome the difficulty, we consider the column-sampling technique and
presume a special form of rotation matrix to facilitate subproblem
decomposition. We test our method on several challenging image datasets and
demonstrate its effectiveness by comparing with state-of-the-art binary coding
techniques.
| no_new_dataset | 0.943191 |
1512.00308 | Luiz Capretz Dr. | Faheem Ahmed, Piers Campbell, Ahmad Jaffar, Luiz Fernando Capretz | Managing Support Requests in Open Source Software Project: The Role of
Online Forums | arXiv admin note: substantial text overlap with arXiv:1507.06927 | 2nd IEEE International Conference on Computer Science and
Information Technology, pp. 590-594, 2009 | 10.1109/ICCSIT.2009.5234491 | null | cs.SE cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The use of free and open source software is gaining momentum due to the ever
increasing availability and use of the Internet. Organizations are also now
adopting open source software, despite some reservations in particular
regarding the provision and availability of support. One of the greatest
concerns about free and open source software is the availability of post
release support and the handling of for support. A common belief is that there
is no appropriate support available for this class of software, while an
alternative argument is that due to the active involvement of Internet users in
online forums, there is in fact a large resource available that communicates
and manages the management of support requests. The research model of this
empirical investigation establishes and studies the relationship between open
source software support requests and online public forums. The results of this
empirical study provide evidence about the realities of support that is present
in open source software projects. We used a dataset consisting of 616 open
source software projects covering a broad range of categories in this
investigation. The results show that online forums play a significant role in
managing support requests in open source software, thus becoming a major source
of assistance in maintenance of the open source projects.
| [
{
"version": "v1",
"created": "Tue, 1 Dec 2015 15:51:24 GMT"
}
] | 2015-12-02T00:00:00 | [
[
"Ahmed",
"Faheem",
""
],
[
"Campbell",
"Piers",
""
],
[
"Jaffar",
"Ahmad",
""
],
[
"Capretz",
"Luiz Fernando",
""
]
] | TITLE: Managing Support Requests in Open Source Software Project: The Role of
Online Forums
ABSTRACT: The use of free and open source software is gaining momentum due to the ever
increasing availability and use of the Internet. Organizations are also now
adopting open source software, despite some reservations in particular
regarding the provision and availability of support. One of the greatest
concerns about free and open source software is the availability of post
release support and the handling of for support. A common belief is that there
is no appropriate support available for this class of software, while an
alternative argument is that due to the active involvement of Internet users in
online forums, there is in fact a large resource available that communicates
and manages the management of support requests. The research model of this
empirical investigation establishes and studies the relationship between open
source software support requests and online public forums. The results of this
empirical study provide evidence about the realities of support that is present
in open source software projects. We used a dataset consisting of 616 open
source software projects covering a broad range of categories in this
investigation. The results show that online forums play a significant role in
managing support requests in open source software, thus becoming a major source
of assistance in maintenance of the open source projects.
| new_dataset | 0.970743 |
1412.5687 | Abhijit Bendale | Abhijit Bendale, Terrance Boult | Towards Open World Recognition | null | IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
(2015) 1893 - 1902 | 10.1109/CVPR.2015.7298799 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the of advent rich classification models and high computational power
visual recognition systems have found many operational applications.
Recognition in the real world poses multiple challenges that are not apparent
in controlled lab environments. The datasets are dynamic and novel categories
must be continuously detected and then added. At prediction time, a trained
system has to deal with myriad unseen categories. Operational systems require
minimum down time, even to learn. To handle these operational issues, we
present the problem of Open World recognition and formally define it. We prove
that thresholding sums of monotonically decreasing functions of distances in
linearly transformed feature space can balance "open space risk" and empirical
risk. Our theory extends existing algorithms for open world recognition. We
present a protocol for evaluation of open world recognition systems. We present
the Nearest Non-Outlier (NNO) algorithm which evolves model efficiently, adding
object categories incrementally while detecting outliers and managing open
space risk. We perform experiments on the ImageNet dataset with 1.2M+ images to
validate the effectiveness of our method on large scale visual recognition
tasks. NNO consistently yields superior results on open world recognition.
| [
{
"version": "v1",
"created": "Thu, 18 Dec 2014 00:07:45 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Bendale",
"Abhijit",
""
],
[
"Boult",
"Terrance",
""
]
] | TITLE: Towards Open World Recognition
ABSTRACT: With the of advent rich classification models and high computational power
visual recognition systems have found many operational applications.
Recognition in the real world poses multiple challenges that are not apparent
in controlled lab environments. The datasets are dynamic and novel categories
must be continuously detected and then added. At prediction time, a trained
system has to deal with myriad unseen categories. Operational systems require
minimum down time, even to learn. To handle these operational issues, we
present the problem of Open World recognition and formally define it. We prove
that thresholding sums of monotonically decreasing functions of distances in
linearly transformed feature space can balance "open space risk" and empirical
risk. Our theory extends existing algorithms for open world recognition. We
present a protocol for evaluation of open world recognition systems. We present
the Nearest Non-Outlier (NNO) algorithm which evolves model efficiently, adding
object categories incrementally while detecting outliers and managing open
space risk. We perform experiments on the ImageNet dataset with 1.2M+ images to
validate the effectiveness of our method on large scale visual recognition
tasks. NNO consistently yields superior results on open world recognition.
| no_new_dataset | 0.945901 |
1503.09129 | Brian Gardner BG | Brian Gardner, Ioana Sporea, Andr\'e Gr\"uning | Encoding Spike Patterns in Multilayer Spiking Neural Networks | 31 pages, 14 figures | null | 10.1162/NECO_a_00790 | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Information encoding in the nervous system is supported through the precise
spike-timings of neurons; however, an understanding of the underlying processes
by which such representations are formed in the first place remains unclear.
Here we examine how networks of spiking neurons can learn to encode for input
patterns using a fully temporal coding scheme. To this end, we introduce a
learning rule for spiking networks containing hidden neurons which optimizes
the likelihood of generating desired output spiking patterns. We show the
proposed learning rule allows for a large number of accurate input-output spike
pattern mappings to be learnt, which outperforms other existing learning rules
for spiking neural networks: both in the number of mappings that can be learnt
as well as the complexity of spike train encodings that can be utilised. The
learning rule is successful even in the presence of input noise, is
demonstrated to solve the linearly non-separable XOR computation and
generalizes well on an example dataset. We further present a biologically
plausible implementation of backpropagated learning in multilayer spiking
networks, and discuss the neural mechanisms that might underlie its function.
Our approach contributes both to a systematic understanding of how pattern
encodings might take place in the nervous system, and a learning rule that
displays strong technical capability.
| [
{
"version": "v1",
"created": "Tue, 31 Mar 2015 17:12:07 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Gardner",
"Brian",
""
],
[
"Sporea",
"Ioana",
""
],
[
"Grüning",
"André",
""
]
] | TITLE: Encoding Spike Patterns in Multilayer Spiking Neural Networks
ABSTRACT: Information encoding in the nervous system is supported through the precise
spike-timings of neurons; however, an understanding of the underlying processes
by which such representations are formed in the first place remains unclear.
Here we examine how networks of spiking neurons can learn to encode for input
patterns using a fully temporal coding scheme. To this end, we introduce a
learning rule for spiking networks containing hidden neurons which optimizes
the likelihood of generating desired output spiking patterns. We show the
proposed learning rule allows for a large number of accurate input-output spike
pattern mappings to be learnt, which outperforms other existing learning rules
for spiking neural networks: both in the number of mappings that can be learnt
as well as the complexity of spike train encodings that can be utilised. The
learning rule is successful even in the presence of input noise, is
demonstrated to solve the linearly non-separable XOR computation and
generalizes well on an example dataset. We further present a biologically
plausible implementation of backpropagated learning in multilayer spiking
networks, and discuss the neural mechanisms that might underlie its function.
Our approach contributes both to a systematic understanding of how pattern
encodings might take place in the nervous system, and a learning rule that
displays strong technical capability.
| no_new_dataset | 0.945399 |
1504.00981 | Wu Ai | Wu Ai and Weisheng Chen | ELM-Based Distributed Cooperative Learning Over Networks | This paper has been withdrawn by the authors due to the incorrect
proof of Theorem 2 | null | null | null | cs.LG math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates distributed cooperative learning algorithms for data
processing in a network setting. Specifically, the extreme learning machine
(ELM) is introduced to train a set of data distributed across several
components, and each component runs a program on a subset of the entire data.
In this scheme, there is no requirement for a fusion center in the network due
to e.g., practical limitations, security, or privacy reasons. We first
reformulate the centralized ELM training problem into a separable form among
nodes with consensus constraints. Then, we solve the equivalent problem using
distributed optimization tools. A new distributed cooperative learning
algorithm based on ELM, called DC-ELM, is proposed. The architecture of this
algorithm differs from that of some existing parallel/distributed ELMs based on
MapReduce or cloud computing. We also present an online version of the proposed
algorithm that can learn data sequentially in a one-by-one or chunk-by-chunk
mode. The novel algorithm is well suited for potential applications such as
artificial intelligence, computational biology, finance, wireless sensor
networks, and so on, involving datasets that are often extremely large,
high-dimensional and located on distributed data sources. We show simulation
results on both synthetic and real-world data sets.
| [
{
"version": "v1",
"created": "Sat, 4 Apr 2015 04:40:48 GMT"
},
{
"version": "v2",
"created": "Mon, 30 Nov 2015 05:48:56 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Ai",
"Wu",
""
],
[
"Chen",
"Weisheng",
""
]
] | TITLE: ELM-Based Distributed Cooperative Learning Over Networks
ABSTRACT: This paper investigates distributed cooperative learning algorithms for data
processing in a network setting. Specifically, the extreme learning machine
(ELM) is introduced to train a set of data distributed across several
components, and each component runs a program on a subset of the entire data.
In this scheme, there is no requirement for a fusion center in the network due
to e.g., practical limitations, security, or privacy reasons. We first
reformulate the centralized ELM training problem into a separable form among
nodes with consensus constraints. Then, we solve the equivalent problem using
distributed optimization tools. A new distributed cooperative learning
algorithm based on ELM, called DC-ELM, is proposed. The architecture of this
algorithm differs from that of some existing parallel/distributed ELMs based on
MapReduce or cloud computing. We also present an online version of the proposed
algorithm that can learn data sequentially in a one-by-one or chunk-by-chunk
mode. The novel algorithm is well suited for potential applications such as
artificial intelligence, computational biology, finance, wireless sensor
networks, and so on, involving datasets that are often extremely large,
high-dimensional and located on distributed data sources. We show simulation
results on both synthetic and real-world data sets.
| no_new_dataset | 0.943034 |
1505.02074 | Mengye Ren | Mengye Ren, Ryan Kiros, Richard Zemel | Exploring Models and Data for Image Question Answering | 12 pages. Conference paper at NIPS 2015 | null | null | null | cs.LG cs.AI cs.CL cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work aims to address the problem of image-based question-answering (QA)
with new models and datasets. In our work, we propose to use neural networks
and visual semantic embeddings, without intermediate stages such as object
detection and image segmentation, to predict answers to simple questions about
images. Our model performs 1.8 times better than the only published results on
an existing image QA dataset. We also present a question generation algorithm
that converts image descriptions, which are widely available, into QA form. We
used this algorithm to produce an order-of-magnitude larger dataset, with more
evenly distributed answers. A suite of baseline results on this new dataset are
also presented.
| [
{
"version": "v1",
"created": "Fri, 8 May 2015 15:59:44 GMT"
},
{
"version": "v2",
"created": "Fri, 19 Jun 2015 19:55:07 GMT"
},
{
"version": "v3",
"created": "Thu, 25 Jun 2015 06:44:44 GMT"
},
{
"version": "v4",
"created": "Sun, 29 Nov 2015 22:45:12 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Ren",
"Mengye",
""
],
[
"Kiros",
"Ryan",
""
],
[
"Zemel",
"Richard",
""
]
] | TITLE: Exploring Models and Data for Image Question Answering
ABSTRACT: This work aims to address the problem of image-based question-answering (QA)
with new models and datasets. In our work, we propose to use neural networks
and visual semantic embeddings, without intermediate stages such as object
detection and image segmentation, to predict answers to simple questions about
images. Our model performs 1.8 times better than the only published results on
an existing image QA dataset. We also present a question generation algorithm
that converts image descriptions, which are widely available, into QA form. We
used this algorithm to produce an order-of-magnitude larger dataset, with more
evenly distributed answers. A suite of baseline results on this new dataset are
also presented.
| new_dataset | 0.952175 |
1506.00262 | Kyle Beauchamp | Kyle A. Beauchamp, Julie M. Behr, Ari\"en S. Rustenburg, Christopher
I. Bayly, Kenneth Kroenlein, John D. Chodera | Towards Automated Benchmarking of Atomistic Forcefields: Neat Liquid
Densities and Static Dielectric Constants from the ThermoML Data Archive | null | J. Phys. Chem. B, 2015, 119 (40), pp 12912-12920 | 10.1021/acs.jpcb.5b06703 | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Atomistic molecular simulations are a powerful way to make quantitative
predictions, but the accuracy of these predictions depends entirely on the
quality of the forcefield employed. While experimental measurements of
fundamental physical properties offer a straightforward approach for evaluating
forcefield quality, the bulk of this information has been tied up in formats
that are not machine-readable. Compiling benchmark datasets of physical
properties from non-machine-readable sources require substantial human effort
and is prone to accumulation of human errors, hindering the development of
reproducible benchmarks of forcefield accuracy. Here, we examine the
feasibility of benchmarking atomistic forcefields against the NIST ThermoML
data archive of physicochemical measurements, which aggregates thousands of
experimental measurements in a portable, machine-readable, self-annotating
format. As a proof of concept, we present a detailed benchmark of the
generalized Amber small molecule forcefield (GAFF) using the AM1-BCC charge
model against measurements (specifically bulk liquid densities and static
dielectric constants at ambient pressure) automatically extracted from the
archive, and discuss the extent of available data. The results of this
benchmark highlight a general problem with fixed-charge forcefields in the
representation low dielectric environments such as those seen in binding
cavities or biological membranes.
| [
{
"version": "v1",
"created": "Sun, 31 May 2015 17:59:03 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Beauchamp",
"Kyle A.",
""
],
[
"Behr",
"Julie M.",
""
],
[
"Rustenburg",
"Ariën S.",
""
],
[
"Bayly",
"Christopher I.",
""
],
[
"Kroenlein",
"Kenneth",
""
],
[
"Chodera",
"John D.",
""
]
] | TITLE: Towards Automated Benchmarking of Atomistic Forcefields: Neat Liquid
Densities and Static Dielectric Constants from the ThermoML Data Archive
ABSTRACT: Atomistic molecular simulations are a powerful way to make quantitative
predictions, but the accuracy of these predictions depends entirely on the
quality of the forcefield employed. While experimental measurements of
fundamental physical properties offer a straightforward approach for evaluating
forcefield quality, the bulk of this information has been tied up in formats
that are not machine-readable. Compiling benchmark datasets of physical
properties from non-machine-readable sources require substantial human effort
and is prone to accumulation of human errors, hindering the development of
reproducible benchmarks of forcefield accuracy. Here, we examine the
feasibility of benchmarking atomistic forcefields against the NIST ThermoML
data archive of physicochemical measurements, which aggregates thousands of
experimental measurements in a portable, machine-readable, self-annotating
format. As a proof of concept, we present a detailed benchmark of the
generalized Amber small molecule forcefield (GAFF) using the AM1-BCC charge
model against measurements (specifically bulk liquid densities and static
dielectric constants at ambient pressure) automatically extracted from the
archive, and discuss the extent of available data. The results of this
benchmark highlight a general problem with fixed-charge forcefields in the
representation low dielectric environments such as those seen in binding
cavities or biological membranes.
| no_new_dataset | 0.931774 |
1511.08827 | Luiz Capretz Dr. | Arif Raza, Luiz Fernando Capretz, Faheem Ahmed | Maintenance Support in Open Source Software Projects | null | null | 10.1109/ICDIM.2013.6694005 | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Easy and mostly free access to the internet has resulted in the growing use
of open source software (OSS). However, it is a common perception that closed
proprietary software is still superior in areas such as software maintenance
and management. The research model of this study establishes a relationship
between maintenance issues (such as user requests and error handling) and
support provided by open source software through project forums, mailing lists
and trackers. To conduct this research, we have used a dataset consisting of
120 open source software projects, covering a wide range of categories. The
results of the study show that project forums and mailing lists play a
significant role in addressing user requests in open source software. However
according to the empirical investigation, it has been explored that trackers
are used as an effective medium for error reporting as well as user requests.
| [
{
"version": "v1",
"created": "Fri, 27 Nov 2015 21:43:39 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Raza",
"Arif",
""
],
[
"Capretz",
"Luiz Fernando",
""
],
[
"Ahmed",
"Faheem",
""
]
] | TITLE: Maintenance Support in Open Source Software Projects
ABSTRACT: Easy and mostly free access to the internet has resulted in the growing use
of open source software (OSS). However, it is a common perception that closed
proprietary software is still superior in areas such as software maintenance
and management. The research model of this study establishes a relationship
between maintenance issues (such as user requests and error handling) and
support provided by open source software through project forums, mailing lists
and trackers. To conduct this research, we have used a dataset consisting of
120 open source software projects, covering a wide range of categories. The
results of the study show that project forums and mailing lists play a
significant role in addressing user requests in open source software. However
according to the empirical investigation, it has been explored that trackers
are used as an effective medium for error reporting as well as user requests.
| new_dataset | 0.960584 |
1511.08899 | Mohamed Moustafa | Mohamed Moustafa | Applying deep learning to classify pornographic images and videos | PSIVT 2015, the final publication is available at link.springer.com | null | null | null | cs.CV cs.MM cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is no secret that pornographic material is now a one-click-away from
everyone, including children and minors. General social media networks are
striving to isolate adult images and videos from normal ones. Intelligent image
analysis methods can help to automatically detect and isolate questionable
images in media. Unfortunately, these methods require vast experience to design
the classifier including one or more of the popular computer vision feature
descriptors. We propose to build a classifier based on one of the recently
flourishing deep learning techniques. Convolutional neural networks contain
many layers for both automatic features extraction and classification. The
benefit is an easier system to build (no need for hand-crafting features and
classifiers). Additionally, our experiments show that it is even more accurate
than the state of the art methods on the most recent benchmark dataset.
| [
{
"version": "v1",
"created": "Sat, 28 Nov 2015 13:55:25 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Moustafa",
"Mohamed",
""
]
] | TITLE: Applying deep learning to classify pornographic images and videos
ABSTRACT: It is no secret that pornographic material is now a one-click-away from
everyone, including children and minors. General social media networks are
striving to isolate adult images and videos from normal ones. Intelligent image
analysis methods can help to automatically detect and isolate questionable
images in media. Unfortunately, these methods require vast experience to design
the classifier including one or more of the popular computer vision feature
descriptors. We propose to build a classifier based on one of the recently
flourishing deep learning techniques. Convolutional neural networks contain
many layers for both automatic features extraction and classification. The
benefit is an easier system to build (no need for hand-crafting features and
classifiers). Additionally, our experiments show that it is even more accurate
than the state of the art methods on the most recent benchmark dataset.
| no_new_dataset | 0.942507 |
1511.08951 | Basura Fernando | Basura Fernando, Efstratios Gavves, Damien Muselet, Tinne Tuytelaars | MidRank: Learning to rank based on subsequences | To appear in ICCV 2015 | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a supervised learning to rank algorithm that effectively orders
images by exploiting the structure in image sequences. Most often in the
supervised learning to rank literature, ranking is approached either by
analyzing pairs of images or by optimizing a list-wise surrogate loss function
on full sequences. In this work we propose MidRank, which learns from
moderately sized sub-sequences instead. These sub-sequences contain useful
structural ranking information that leads to better learnability during
training and better generalization during testing. By exploiting sub-sequences,
the proposed MidRank improves ranking accuracy considerably on an extensive
array of image ranking applications and datasets.
| [
{
"version": "v1",
"created": "Sun, 29 Nov 2015 00:47:19 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Fernando",
"Basura",
""
],
[
"Gavves",
"Efstratios",
""
],
[
"Muselet",
"Damien",
""
],
[
"Tuytelaars",
"Tinne",
""
]
] | TITLE: MidRank: Learning to rank based on subsequences
ABSTRACT: We present a supervised learning to rank algorithm that effectively orders
images by exploiting the structure in image sequences. Most often in the
supervised learning to rank literature, ranking is approached either by
analyzing pairs of images or by optimizing a list-wise surrogate loss function
on full sequences. In this work we propose MidRank, which learns from
moderately sized sub-sequences instead. These sub-sequences contain useful
structural ranking information that leads to better learnability during
training and better generalization during testing. By exploiting sub-sequences,
the proposed MidRank improves ranking accuracy considerably on an extensive
array of image ranking applications and datasets.
| no_new_dataset | 0.951997 |
1511.09066 | Richard McClatchey | Kamran Munir, Khawar Hasham Ahmad and Richard McClatchey | Development of a Large-scale Neuroimages and Clinical Variables Data
Atlas in the neuGRID4You (N4U) project | 35 pages, 15 figures, Journal of Biomedical Informatics, 2015 | null | 10.1016/j.jbi.2015.08.004 | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Exceptional growth in the availability of large-scale clinical imaging
datasets has led to the development of computational infrastructures offering
scientists access to image repositories and associated clinical variables data.
The EU FP7 neuGRID and its follow on neuGRID4You (N4U) project is a leading
e-Infrastructure where neuroscientists can find core services and resources for
brain image analysis. The core component of this e-Infrastructure is the N4U
Virtual Laboratory, which offers an easy access for neuroscientists to a wide
range of datasets and algorithms, pipelines, computational resources, services,
and associated support services. The foundation of this virtual laboratory is a
massive data store plus information services called the Data Atlas that stores
datasets, clinical study data, data dictionaries, algorithm/pipeline
definitions, and provides interfaces for parameterised querying so that
neuroscientists can perform analyses on required datasets. This paper presents
the overall design and development of the Data Atlas, its associated datasets
and indexing and a set of retrieval services that originated from the
development of the N4U Virtual Laboratory in the EU FP7 N4U project in the
light of user requirements.
| [
{
"version": "v1",
"created": "Sun, 29 Nov 2015 19:14:53 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Munir",
"Kamran",
""
],
[
"Ahmad",
"Khawar Hasham",
""
],
[
"McClatchey",
"Richard",
""
]
] | TITLE: Development of a Large-scale Neuroimages and Clinical Variables Data
Atlas in the neuGRID4You (N4U) project
ABSTRACT: Exceptional growth in the availability of large-scale clinical imaging
datasets has led to the development of computational infrastructures offering
scientists access to image repositories and associated clinical variables data.
The EU FP7 neuGRID and its follow on neuGRID4You (N4U) project is a leading
e-Infrastructure where neuroscientists can find core services and resources for
brain image analysis. The core component of this e-Infrastructure is the N4U
Virtual Laboratory, which offers an easy access for neuroscientists to a wide
range of datasets and algorithms, pipelines, computational resources, services,
and associated support services. The foundation of this virtual laboratory is a
massive data store plus information services called the Data Atlas that stores
datasets, clinical study data, data dictionaries, algorithm/pipeline
definitions, and provides interfaces for parameterised querying so that
neuroscientists can perform analyses on required datasets. This paper presents
the overall design and development of the Data Atlas, its associated datasets
and indexing and a set of retrieval services that originated from the
development of the N4U Virtual Laboratory in the EU FP7 N4U project in the
light of user requirements.
| no_new_dataset | 0.951278 |
1511.09067 | Mohamed Elawady | Mohamed Elawady | Sparse Coral Classification Using Deep Convolutional Neural Networks | Thesis Submitted for the Degree of MSc Erasmus Mundus in Vision and
Robotics (VIBOT 2014) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Autonomous repair of deep-sea coral reefs is a recent proposed idea to
support the oceans ecosystem in which is vital for commercial fishing, tourism
and other species. This idea can be operated through using many small
autonomous underwater vehicles (AUVs) and swarm intelligence techniques to
locate and replace chunks of coral which have been broken off, thus enabling
re-growth and maintaining the habitat. The aim of this project is developing
machine vision algorithms to enable an underwater robot to locate a coral reef
and a chunk of coral on the seabed and prompt the robot to pick it up. Although
there is no literature on this particular problem, related work on fish
counting may give some insight into the problem. The technical challenges are
principally due to the potential lack of clarity of the water and platform
stabilization as well as spurious artifacts (rocks, fish, and crabs). We
present an efficient sparse classification for coral species using supervised
deep learning method called Convolutional Neural Networks (CNNs). We compute
Weber Local Descriptor (WLD), Phase Congruency (PC), and Zero Component
Analysis (ZCA) Whitening to extract shape and texture feature descriptors,
which are employed to be supplementary channels (feature-based maps) besides
basic spatial color channels (spatial-based maps) of coral input image, we also
experiment state-of-art preprocessing underwater algorithms for image
enhancement and color normalization and color conversion adjustment. Our
proposed coral classification method is developed under MATLAB platform, and
evaluated by two different coral datasets (University of California San Diego's
Moorea Labeled Corals, and Heriot-Watt University's Atlantic Deep Sea).
| [
{
"version": "v1",
"created": "Sun, 29 Nov 2015 19:18:36 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Elawady",
"Mohamed",
""
]
] | TITLE: Sparse Coral Classification Using Deep Convolutional Neural Networks
ABSTRACT: Autonomous repair of deep-sea coral reefs is a recent proposed idea to
support the oceans ecosystem in which is vital for commercial fishing, tourism
and other species. This idea can be operated through using many small
autonomous underwater vehicles (AUVs) and swarm intelligence techniques to
locate and replace chunks of coral which have been broken off, thus enabling
re-growth and maintaining the habitat. The aim of this project is developing
machine vision algorithms to enable an underwater robot to locate a coral reef
and a chunk of coral on the seabed and prompt the robot to pick it up. Although
there is no literature on this particular problem, related work on fish
counting may give some insight into the problem. The technical challenges are
principally due to the potential lack of clarity of the water and platform
stabilization as well as spurious artifacts (rocks, fish, and crabs). We
present an efficient sparse classification for coral species using supervised
deep learning method called Convolutional Neural Networks (CNNs). We compute
Weber Local Descriptor (WLD), Phase Congruency (PC), and Zero Component
Analysis (ZCA) Whitening to extract shape and texture feature descriptors,
which are employed to be supplementary channels (feature-based maps) besides
basic spatial color channels (spatial-based maps) of coral input image, we also
experiment state-of-art preprocessing underwater algorithms for image
enhancement and color normalization and color conversion adjustment. Our
proposed coral classification method is developed under MATLAB platform, and
evaluated by two different coral datasets (University of California San Diego's
Moorea Labeled Corals, and Heriot-Watt University's Atlantic Deep Sea).
| no_new_dataset | 0.954308 |
1511.09107 | K. Ch. Chatzisavvas | Panagiotis Stalidis, Maria Giatsoglou, Konstantinos Diamantaras,
George Sarigiannidis, Konstantinos Ch. Chatzisavvas | Machine Learning Sentiment Prediction based on Hybrid Document
Representation | null | null | null | null | cs.CL cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated sentiment analysis and opinion mining is a complex process
concerning the extraction of useful subjective information from text. The
explosion of user generated content on the Web, especially the fact that
millions of users, on a daily basis, express their opinions on products and
services to blogs, wikis, social networks, message boards, etc., render the
reliable, automated export of sentiments and opinions from unstructured text
crucial for several commercial applications. In this paper, we present a novel
hybrid vectorization approach for textual resources that combines a weighted
variant of the popular Word2Vec representation (based on Term Frequency-Inverse
Document Frequency) representation and with a Bag- of-Words representation and
a vector of lexicon-based sentiment values. The proposed text representation
approach is assessed through the application of several machine learning
classification algorithms on a dataset that is used extensively in literature
for sentiment detection. The classification accuracy derived through the
proposed hybrid vectorization approach is higher than when its individual
components are used for text represenation, and comparable with
state-of-the-art sentiment detection methodologies.
| [
{
"version": "v1",
"created": "Sun, 29 Nov 2015 22:41:43 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Stalidis",
"Panagiotis",
""
],
[
"Giatsoglou",
"Maria",
""
],
[
"Diamantaras",
"Konstantinos",
""
],
[
"Sarigiannidis",
"George",
""
],
[
"Chatzisavvas",
"Konstantinos Ch.",
""
]
] | TITLE: Machine Learning Sentiment Prediction based on Hybrid Document
Representation
ABSTRACT: Automated sentiment analysis and opinion mining is a complex process
concerning the extraction of useful subjective information from text. The
explosion of user generated content on the Web, especially the fact that
millions of users, on a daily basis, express their opinions on products and
services to blogs, wikis, social networks, message boards, etc., render the
reliable, automated export of sentiments and opinions from unstructured text
crucial for several commercial applications. In this paper, we present a novel
hybrid vectorization approach for textual resources that combines a weighted
variant of the popular Word2Vec representation (based on Term Frequency-Inverse
Document Frequency) representation and with a Bag- of-Words representation and
a vector of lexicon-based sentiment values. The proposed text representation
approach is assessed through the application of several machine learning
classification algorithms on a dataset that is used extensively in literature
for sentiment detection. The classification accuracy derived through the
proposed hybrid vectorization approach is higher than when its individual
components are used for text represenation, and comparable with
state-of-the-art sentiment detection methodologies.
| no_new_dataset | 0.947186 |
1511.09134 | Liu Weiyi | Liu Weiyi, Chen Lingli, Hu Guangmin | Mining Essential Relationships under Multiplex Networks | 5 pages, 6 figures | null | null | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In big data times, massive datasets often carry different relationships among
the same group of nodes, analyzing on these heterogeneous relationships may
give us a window to peek the essential relationships among nodes. In this
paper, first of all we propose a new metric "similarity rate" in order to
capture the changing rate of similarities between node-pairs though all
networks; secondly, we try to use this new metric to uncover essential
relationships between node-pairs which essential relationships are often hidden
and hard to get. From experiments study of Indonesian Terrorists dataset, this
new metric similarity rate function well for giving us a way to uncover
essential relationships from lots of appearances.
| [
{
"version": "v1",
"created": "Mon, 30 Nov 2015 02:12:13 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Weiyi",
"Liu",
""
],
[
"Lingli",
"Chen",
""
],
[
"Guangmin",
"Hu",
""
]
] | TITLE: Mining Essential Relationships under Multiplex Networks
ABSTRACT: In big data times, massive datasets often carry different relationships among
the same group of nodes, analyzing on these heterogeneous relationships may
give us a window to peek the essential relationships among nodes. In this
paper, first of all we propose a new metric "similarity rate" in order to
capture the changing rate of similarities between node-pairs though all
networks; secondly, we try to use this new metric to uncover essential
relationships between node-pairs which essential relationships are often hidden
and hard to get. From experiments study of Indonesian Terrorists dataset, this
new metric similarity rate function well for giving us a way to uncover
essential relationships from lots of appearances.
| no_new_dataset | 0.932913 |
1511.09150 | Rahul Rama Varior Mr. | Rahul Rama Varior, Gang Wang | Hierarchical Invariant Feature Learning with Marginalization for Person
Re-Identification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper addresses the problem of matching pedestrians across multiple
camera views, known as person re-identification. Variations in lighting
conditions, environment and pose changes across camera views make
re-identification a challenging problem. Previous methods address these
challenges by designing specific features or by learning a distance function.
We propose a hierarchical feature learning framework that learns invariant
representations from labeled image pairs. A mapping is learned such that the
extracted features are invariant for images belonging to same individual across
views. To learn robust representations and to achieve better generalization to
unseen data, the system has to be trained with a large amount of data.
Critically, most of the person re-identification datasets are small. Manually
augmenting the dataset by partial corruption of input data introduces
additional computational burden as it requires several training epochs to
converge. We propose a hierarchical network which incorporates a
marginalization technique that can reap the benefits of training on large
datasets without explicit augmentation. We compare our approach with several
baseline algorithms as well as popular linear and non-linear metric learning
algorithms and demonstrate improved performance on challenging publicly
available datasets, VIPeR, CUHK01, CAVIAR4REID and iLIDS. Our approach also
achieves the stateof-the-art results on these datasets.
| [
{
"version": "v1",
"created": "Mon, 30 Nov 2015 04:05:21 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Varior",
"Rahul Rama",
""
],
[
"Wang",
"Gang",
""
]
] | TITLE: Hierarchical Invariant Feature Learning with Marginalization for Person
Re-Identification
ABSTRACT: This paper addresses the problem of matching pedestrians across multiple
camera views, known as person re-identification. Variations in lighting
conditions, environment and pose changes across camera views make
re-identification a challenging problem. Previous methods address these
challenges by designing specific features or by learning a distance function.
We propose a hierarchical feature learning framework that learns invariant
representations from labeled image pairs. A mapping is learned such that the
extracted features are invariant for images belonging to same individual across
views. To learn robust representations and to achieve better generalization to
unseen data, the system has to be trained with a large amount of data.
Critically, most of the person re-identification datasets are small. Manually
augmenting the dataset by partial corruption of input data introduces
additional computational burden as it requires several training epochs to
converge. We propose a hierarchical network which incorporates a
marginalization technique that can reap the benefits of training on large
datasets without explicit augmentation. We compare our approach with several
baseline algorithms as well as popular linear and non-linear metric learning
algorithms and demonstrate improved performance on challenging publicly
available datasets, VIPeR, CUHK01, CAVIAR4REID and iLIDS. Our approach also
achieves the stateof-the-art results on these datasets.
| no_new_dataset | 0.9463 |
1511.09159 | Yangyang Xu | Yangyang Xu, Ioannis Akrotirianakis, Amit Chakraborty | Proximal gradient method for huberized support vector machine | in Pattern analysis and application, 2015 | null | 10.1007/s10044-015-0485-z | null | stat.ML cs.LG cs.NA math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Support Vector Machine (SVM) has been used in a wide variety of
classification problems. The original SVM uses the hinge loss function, which
is non-differentiable and makes the problem difficult to solve in particular
for regularized SVMs, such as with $\ell_1$-regularization. This paper
considers the Huberized SVM (HSVM), which uses a differentiable approximation
of the hinge loss function. We first explore the use of the Proximal Gradient
(PG) method to solving binary-class HSVM (B-HSVM) and then generalize it to
multi-class HSVM (M-HSVM). Under strong convexity assumptions, we show that our
algorithm converges linearly. In addition, we give a finite convergence result
about the support of the solution, based on which we further accelerate the
algorithm by a two-stage method. We present extensive numerical experiments on
both synthetic and real datasets which demonstrate the superiority of our
methods over some state-of-the-art methods for both binary- and multi-class
SVMs.
| [
{
"version": "v1",
"created": "Mon, 30 Nov 2015 05:02:02 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Xu",
"Yangyang",
""
],
[
"Akrotirianakis",
"Ioannis",
""
],
[
"Chakraborty",
"Amit",
""
]
] | TITLE: Proximal gradient method for huberized support vector machine
ABSTRACT: The Support Vector Machine (SVM) has been used in a wide variety of
classification problems. The original SVM uses the hinge loss function, which
is non-differentiable and makes the problem difficult to solve in particular
for regularized SVMs, such as with $\ell_1$-regularization. This paper
considers the Huberized SVM (HSVM), which uses a differentiable approximation
of the hinge loss function. We first explore the use of the Proximal Gradient
(PG) method to solving binary-class HSVM (B-HSVM) and then generalize it to
multi-class HSVM (M-HSVM). Under strong convexity assumptions, we show that our
algorithm converges linearly. In addition, we give a finite convergence result
about the support of the solution, based on which we further accelerate the
algorithm by a two-stage method. We present extensive numerical experiments on
both synthetic and real datasets which demonstrate the superiority of our
methods over some state-of-the-art methods for both binary- and multi-class
SVMs.
| no_new_dataset | 0.947914 |
1511.09209 | Zongyuan Ge | ZongYuan Ge and Alex Bewley and Christopher McCool and Ben Upcroft and
Peter Corke and Conrad Sanderson | Fine-Grained Classification via Mixture of Deep Convolutional Neural
Networks | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a novel deep convolutional neural network (DCNN) system for
fine-grained image classification, called a mixture of DCNNs (MixDCNN). The
fine-grained image classification problem is characterised by large intra-class
variations and small inter-class variations. To overcome these problems our
proposed MixDCNN system partitions images into K subsets of similar images and
learns an expert DCNN for each subset. The output from each of the K DCNNs is
combined to form a single classification decision. In contrast to previous
techniques, we provide a formulation to perform joint end-to-end training of
the K DCNNs simultaneously. Extensive experiments, on three datasets using two
network structures (AlexNet and GoogLeNet), show that the proposed MixDCNN
system consistently outperforms other methods. It provides a relative
improvement of 12.7% and achieves state-of-the-art results on two datasets.
| [
{
"version": "v1",
"created": "Mon, 30 Nov 2015 09:14:10 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Ge",
"ZongYuan",
""
],
[
"Bewley",
"Alex",
""
],
[
"McCool",
"Christopher",
""
],
[
"Upcroft",
"Ben",
""
],
[
"Corke",
"Peter",
""
],
[
"Sanderson",
"Conrad",
""
]
] | TITLE: Fine-Grained Classification via Mixture of Deep Convolutional Neural
Networks
ABSTRACT: We present a novel deep convolutional neural network (DCNN) system for
fine-grained image classification, called a mixture of DCNNs (MixDCNN). The
fine-grained image classification problem is characterised by large intra-class
variations and small inter-class variations. To overcome these problems our
proposed MixDCNN system partitions images into K subsets of similar images and
learns an expert DCNN for each subset. The output from each of the K DCNNs is
combined to form a single classification decision. In contrast to previous
techniques, we provide a formulation to perform joint end-to-end training of
the K DCNNs simultaneously. Extensive experiments, on three datasets using two
network structures (AlexNet and GoogLeNet), show that the proposed MixDCNN
system consistently outperforms other methods. It provides a relative
improvement of 12.7% and achieves state-of-the-art results on two datasets.
| no_new_dataset | 0.948298 |
1511.09460 | Snigdha Chaturvedi | Snigdha Chaturvedi, Dan Goldwasser, Hal Daume III | Ask, and shall you receive?: Understanding Desire Fulfillment in Natural
Language Text | null | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ability to comprehend wishes or desires and their fulfillment is
important to Natural Language Understanding. This paper introduces the task of
identifying if a desire expressed by a subject in a given short piece of text
was fulfilled. We propose various unstructured and structured models that
capture fulfillment cues such as the subject's emotional state and actions. Our
experiments with two different datasets demonstrate the importance of
understanding the narrative and discourse structure to address this task.
| [
{
"version": "v1",
"created": "Mon, 30 Nov 2015 20:37:03 GMT"
}
] | 2015-12-01T00:00:00 | [
[
"Chaturvedi",
"Snigdha",
""
],
[
"Goldwasser",
"Dan",
""
],
[
"Daume",
"Hal",
"III"
]
] | TITLE: Ask, and shall you receive?: Understanding Desire Fulfillment in Natural
Language Text
ABSTRACT: The ability to comprehend wishes or desires and their fulfillment is
important to Natural Language Understanding. This paper introduces the task of
identifying if a desire expressed by a subject in a given short piece of text
was fulfilled. We propose various unstructured and structured models that
capture fulfillment cues such as the subject's emotional state and actions. Our
experiments with two different datasets demonstrate the importance of
understanding the narrative and discourse structure to address this task.
| no_new_dataset | 0.945147 |
1505.00853 | Bing Xu | Bing Xu, Naiyan Wang, Tianqi Chen, Mu Li | Empirical Evaluation of Rectified Activations in Convolutional Network | null | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we investigate the performance of different types of rectified
activation functions in convolutional neural network: standard rectified linear
unit (ReLU), leaky rectified linear unit (Leaky ReLU), parametric rectified
linear unit (PReLU) and a new randomized leaky rectified linear units (RReLU).
We evaluate these activation function on standard image classification task.
Our experiments suggest that incorporating a non-zero slope for negative part
in rectified activation units could consistently improve the results. Thus our
findings are negative on the common belief that sparsity is the key of good
performance in ReLU. Moreover, on small scale dataset, using deterministic
negative slope or learning it are both prone to overfitting. They are not as
effective as using their randomized counterpart. By using RReLU, we achieved
75.68\% accuracy on CIFAR-100 test set without multiple test or ensemble.
| [
{
"version": "v1",
"created": "Tue, 5 May 2015 01:16:39 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Nov 2015 06:58:14 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Xu",
"Bing",
""
],
[
"Wang",
"Naiyan",
""
],
[
"Chen",
"Tianqi",
""
],
[
"Li",
"Mu",
""
]
] | TITLE: Empirical Evaluation of Rectified Activations in Convolutional Network
ABSTRACT: In this paper we investigate the performance of different types of rectified
activation functions in convolutional neural network: standard rectified linear
unit (ReLU), leaky rectified linear unit (Leaky ReLU), parametric rectified
linear unit (PReLU) and a new randomized leaky rectified linear units (RReLU).
We evaluate these activation function on standard image classification task.
Our experiments suggest that incorporating a non-zero slope for negative part
in rectified activation units could consistently improve the results. Thus our
findings are negative on the common belief that sparsity is the key of good
performance in ReLU. Moreover, on small scale dataset, using deterministic
negative slope or learning it are both prone to overfitting. They are not as
effective as using their randomized counterpart. By using RReLU, we achieved
75.68\% accuracy on CIFAR-100 test set without multiple test or ensemble.
| no_new_dataset | 0.954351 |
1511.07041 | Ankur Handa | Ankur Handa, Viorica Patraucean, Vijay Badrinarayanan, Simon Stent,
Roberto Cipolla | SceneNet: Understanding Real World Indoor Scenes With Synthetic Data | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Scene understanding is a prerequisite to many high level tasks for any
automated intelligent machine operating in real world environments. Recent
attempts with supervised learning have shown promise in this direction but also
highlighted the need for enormous quantity of supervised data --- performance
increases in proportion to the amount of data used. However, this quickly
becomes prohibitive when considering the manual labour needed to collect such
data. In this work, we focus our attention on depth based semantic per-pixel
labelling as a scene understanding problem and show the potential of computer
graphics to generate virtually unlimited labelled data from synthetic 3D
scenes. By carefully synthesizing training data with appropriate noise models
we show comparable performance to state-of-the-art RGBD systems on NYUv2
dataset despite using only depth data as input and set a benchmark on
depth-based segmentation on SUN RGB-D dataset. Additionally, we offer a route
to generating synthesized frame or video data, and understanding of different
factors influencing performance gains.
| [
{
"version": "v1",
"created": "Sun, 22 Nov 2015 17:59:49 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Nov 2015 22:09:09 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Handa",
"Ankur",
""
],
[
"Patraucean",
"Viorica",
""
],
[
"Badrinarayanan",
"Vijay",
""
],
[
"Stent",
"Simon",
""
],
[
"Cipolla",
"Roberto",
""
]
] | TITLE: SceneNet: Understanding Real World Indoor Scenes With Synthetic Data
ABSTRACT: Scene understanding is a prerequisite to many high level tasks for any
automated intelligent machine operating in real world environments. Recent
attempts with supervised learning have shown promise in this direction but also
highlighted the need for enormous quantity of supervised data --- performance
increases in proportion to the amount of data used. However, this quickly
becomes prohibitive when considering the manual labour needed to collect such
data. In this work, we focus our attention on depth based semantic per-pixel
labelling as a scene understanding problem and show the potential of computer
graphics to generate virtually unlimited labelled data from synthetic 3D
scenes. By carefully synthesizing training data with appropriate noise models
we show comparable performance to state-of-the-art RGBD systems on NYUv2
dataset despite using only depth data as input and set a benchmark on
depth-based segmentation on SUN RGB-D dataset. Additionally, we offer a route
to generating synthesized frame or video data, and understanding of different
factors influencing performance gains.
| no_new_dataset | 0.949856 |
1511.08350 | Amina Kemmar | Amina Kemmar and Samir Loudni and Yahia Lebbah and Patrice Boizumault
and Thierry Charnois | A global Constraint for mining Sequential Patterns with GAP constraint | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sequential pattern mining (SPM) under gap constraint is a challenging task.
Many efficient specialized methods have been developed but they are all
suffering from a lack of genericity. The Constraint Programming (CP) approaches
are not so effective because of the size of their encodings. In[7], we have
proposed the global constraint Prefix-Projection for SPM which remedies to this
drawback. However, this global constraint cannot be directly extended to
support gap constraint. In this paper, we propose the global constraint GAP-SEQ
enabling to handle SPM with or without gap constraint. GAP-SEQ relies on the
principle of right pattern extensions. Experiments show that our approach
clearly outperforms both CP approaches and the state-of-the-art cSpade method
on large datasets.
| [
{
"version": "v1",
"created": "Thu, 26 Nov 2015 10:45:34 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Kemmar",
"Amina",
""
],
[
"Loudni",
"Samir",
""
],
[
"Lebbah",
"Yahia",
""
],
[
"Boizumault",
"Patrice",
""
],
[
"Charnois",
"Thierry",
""
]
] | TITLE: A global Constraint for mining Sequential Patterns with GAP constraint
ABSTRACT: Sequential pattern mining (SPM) under gap constraint is a challenging task.
Many efficient specialized methods have been developed but they are all
suffering from a lack of genericity. The Constraint Programming (CP) approaches
are not so effective because of the size of their encodings. In[7], we have
proposed the global constraint Prefix-Projection for SPM which remedies to this
drawback. However, this global constraint cannot be directly extended to
support gap constraint. In this paper, we propose the global constraint GAP-SEQ
enabling to handle SPM with or without gap constraint. GAP-SEQ relies on the
principle of right pattern extensions. Experiments show that our approach
clearly outperforms both CP approaches and the state-of-the-art cSpade method
on large datasets.
| no_new_dataset | 0.949902 |
1511.08411 | Mostafa Bayomi | Mostafa Bayomi, Killian Levacher, M. Rami Ghorab, S\'eamus Lawless | OntoSeg: a Novel Approach to Text Segmentation using Ontological
Similarity | 10 pages, IEEE ICDMW 2015 (SENTIRE Workshop) | null | 10.1109/ICDMW.2015.6 | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text segmentation (TS) aims at dividing long text into coherent segments
which reflect the subtopic structure of the text. It is beneficial to many
natural language processing tasks, such as Information Retrieval (IR) and
document summarisation. Current approaches to text segmentation are similar in
that they all use word-frequency metrics to measure the similarity between two
regions of text, so that a document is segmented based on the lexical cohesion
between its words. Various NLP tasks are now moving towards the semantic web
and ontologies, such as ontology-based IR systems, to capture the
conceptualizations associated with user needs and contents. Text segmentation
based on lexical cohesion between words is hence not sufficient anymore for
such tasks. This paper proposes OntoSeg, a novel approach to text segmentation
based on the ontological similarity between text blocks. The proposed method
uses ontological similarity to explore conceptual relations between text
segments and a Hierarchical Agglomerative Clustering (HAC) algorithm to
represent the text as a tree-like hierarchy that is conceptually structured.
The rich structure of the created tree further allows the segmentation of text
in a linear fashion at various levels of granularity. The proposed method was
evaluated on a wellknown dataset, and the results show that using ontological
similarity in text segmentation is very promising. Also we enhance the proposed
method by combining ontological similarity with lexical similarity and the
results show an enhancement of the segmentation quality.
| [
{
"version": "v1",
"created": "Thu, 26 Nov 2015 15:10:18 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Bayomi",
"Mostafa",
""
],
[
"Levacher",
"Killian",
""
],
[
"Ghorab",
"M. Rami",
""
],
[
"Lawless",
"Séamus",
""
]
] | TITLE: OntoSeg: a Novel Approach to Text Segmentation using Ontological
Similarity
ABSTRACT: Text segmentation (TS) aims at dividing long text into coherent segments
which reflect the subtopic structure of the text. It is beneficial to many
natural language processing tasks, such as Information Retrieval (IR) and
document summarisation. Current approaches to text segmentation are similar in
that they all use word-frequency metrics to measure the similarity between two
regions of text, so that a document is segmented based on the lexical cohesion
between its words. Various NLP tasks are now moving towards the semantic web
and ontologies, such as ontology-based IR systems, to capture the
conceptualizations associated with user needs and contents. Text segmentation
based on lexical cohesion between words is hence not sufficient anymore for
such tasks. This paper proposes OntoSeg, a novel approach to text segmentation
based on the ontological similarity between text blocks. The proposed method
uses ontological similarity to explore conceptual relations between text
segments and a Hierarchical Agglomerative Clustering (HAC) algorithm to
represent the text as a tree-like hierarchy that is conceptually structured.
The rich structure of the created tree further allows the segmentation of text
in a linear fashion at various levels of granularity. The proposed method was
evaluated on a wellknown dataset, and the results show that using ontological
similarity in text segmentation is very promising. Also we enhance the proposed
method by combining ontological similarity with lexical similarity and the
results show an enhancement of the segmentation quality.
| no_new_dataset | 0.945298 |
1511.08417 | Ziqiang Cao | Ziqiang Cao, Chengyao Chen, Wenjie Li, Sujian Li, Furu Wei, Ming Zhou | TGSum: Build Tweet Guided Multi-Document Summarization Dataset | 7 pages, 1 figure in AAAI 2016 | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The development of summarization research has been significantly hampered by
the costly acquisition of reference summaries. This paper proposes an effective
way to automatically collect large scales of news-related multi-document
summaries with reference to social media's reactions. We utilize two types of
social labels in tweets, i.e., hashtags and hyper-links. Hashtags are used to
cluster documents into different topic sets. Also, a tweet with a hyper-link
often highlights certain key points of the corresponding document. We
synthesize a linked document cluster to form a reference summary which can
cover most key points. To this aim, we adopt the ROUGE metrics to measure the
coverage ratio, and develop an Integer Linear Programming solution to discover
the sentence set reaching the upper bound of ROUGE. Since we allow summary
sentences to be selected from both documents and high-quality tweets, the
generated reference summaries could be abstractive. Both informativeness and
readability of the collected summaries are verified by manual judgment. In
addition, we train a Support Vector Regression summarizer on DUC generic
multi-document summarization benchmarks. With the collected data as extra
training resource, the performance of the summarizer improves a lot on all the
test sets. We release this dataset for further research.
| [
{
"version": "v1",
"created": "Thu, 26 Nov 2015 15:22:54 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Cao",
"Ziqiang",
""
],
[
"Chen",
"Chengyao",
""
],
[
"Li",
"Wenjie",
""
],
[
"Li",
"Sujian",
""
],
[
"Wei",
"Furu",
""
],
[
"Zhou",
"Ming",
""
]
] | TITLE: TGSum: Build Tweet Guided Multi-Document Summarization Dataset
ABSTRACT: The development of summarization research has been significantly hampered by
the costly acquisition of reference summaries. This paper proposes an effective
way to automatically collect large scales of news-related multi-document
summaries with reference to social media's reactions. We utilize two types of
social labels in tweets, i.e., hashtags and hyper-links. Hashtags are used to
cluster documents into different topic sets. Also, a tweet with a hyper-link
often highlights certain key points of the corresponding document. We
synthesize a linked document cluster to form a reference summary which can
cover most key points. To this aim, we adopt the ROUGE metrics to measure the
coverage ratio, and develop an Integer Linear Programming solution to discover
the sentence set reaching the upper bound of ROUGE. Since we allow summary
sentences to be selected from both documents and high-quality tweets, the
generated reference summaries could be abstractive. Both informativeness and
readability of the collected summaries are verified by manual judgment. In
addition, we train a Support Vector Regression summarizer on DUC generic
multi-document summarization benchmarks. With the collected data as extra
training resource, the performance of the summarizer improves a lot on all the
test sets. We release this dataset for further research.
| new_dataset | 0.957358 |
1511.08446 | Xu Jia | Amir Ghodrati and Xu Jia and Marco Pedersoli and Tinne Tuytelaars | Towards Automatic Image Editing: Learning to See another You | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning the distribution of images in order to generate new samples is a
challenging task due to the high dimensionality of the data and the highly
non-linear relations that are involved. Nevertheless, some promising results
have been reported in the literature recently,building on deep network
architectures. In this work, we zoom in on a specific type of image generation:
given an image and knowing the category of objects it belongs to (e.g. faces),
our goal is to generate a similar and plausible image, but with some altered
attributes. This is particularly challenging, as the model needs to learn to
disentangle the effect of each attribute and to apply a desired attribute
change to a given input image, while keeping the other attributes and overall
object appearance intact. To this end, we learn a convolutional network, where
the desired attribute information is encoded then merged with the encoded image
at feature map level. We show promising results, both qualitatively as well as
quantitatively, in the context of a retrieval experiment, on two face datasets
(MultiPie and CAS-PEAL-R1).
| [
{
"version": "v1",
"created": "Thu, 26 Nov 2015 16:33:10 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Ghodrati",
"Amir",
""
],
[
"Jia",
"Xu",
""
],
[
"Pedersoli",
"Marco",
""
],
[
"Tuytelaars",
"Tinne",
""
]
] | TITLE: Towards Automatic Image Editing: Learning to See another You
ABSTRACT: Learning the distribution of images in order to generate new samples is a
challenging task due to the high dimensionality of the data and the highly
non-linear relations that are involved. Nevertheless, some promising results
have been reported in the literature recently,building on deep network
architectures. In this work, we zoom in on a specific type of image generation:
given an image and knowing the category of objects it belongs to (e.g. faces),
our goal is to generate a similar and plausible image, but with some altered
attributes. This is particularly challenging, as the model needs to learn to
disentangle the effect of each attribute and to apply a desired attribute
change to a given input image, while keeping the other attributes and overall
object appearance intact. To this end, we learn a convolutional network, where
the desired attribute information is encoded then merged with the encoded image
at feature map level. We show promising results, both qualitatively as well as
quantitatively, in the context of a retrieval experiment, on two face datasets
(MultiPie and CAS-PEAL-R1).
| no_new_dataset | 0.948298 |
1511.08707 | Md Azharuddin | Tripti Tanaya Tejaswi, Md Azharuddin, P. K. Jana | A GA based approach for task scheduling in multi-cloud environment | null | null | null | null | cs.DC | http://creativecommons.org/publicdomain/zero/1.0/ | In multi-cloud environment, task scheduling has attracted a lot of attention
due to NP-Complete nature of the problem. Moreover, it is very challenging due
to heterogeneity of the cloud resources with varying capacities and
functionalities. Therefore, minimizing the makespan for task scheduling is a
challenging issue. In this paper, we propose a genetic algorithm (GA) based
approach for solving task scheduling problem. The algorithm is described with
innovative idea of fitness function derivation and mutation. The proposed
algorithm is exposed to rigorous testing using various benchmark datasets and
its performance is evaluated in terms of total makespan.
| [
{
"version": "v1",
"created": "Fri, 27 Nov 2015 15:30:13 GMT"
}
] | 2015-11-30T00:00:00 | [
[
"Tejaswi",
"Tripti Tanaya",
""
],
[
"Azharuddin",
"Md",
""
],
[
"Jana",
"P. K.",
""
]
] | TITLE: A GA based approach for task scheduling in multi-cloud environment
ABSTRACT: In multi-cloud environment, task scheduling has attracted a lot of attention
due to NP-Complete nature of the problem. Moreover, it is very challenging due
to heterogeneity of the cloud resources with varying capacities and
functionalities. Therefore, minimizing the makespan for task scheduling is a
challenging issue. In this paper, we propose a genetic algorithm (GA) based
approach for solving task scheduling problem. The algorithm is described with
innovative idea of fitness function derivation and mutation. The proposed
algorithm is exposed to rigorous testing using various benchmark datasets and
its performance is evaluated in terms of total makespan.
| no_new_dataset | 0.952131 |
1508.03881 | Fangting Xia | Fangting Xia, Jun Zhu, Peng Wang, Alan Yuille | Pose-Guided Human Parsing with Deep Learned Features | 12 pages, 10 figures, a shortened version of this paper was accepted
by AAAI 2016 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Parsing human body into semantic regions is crucial to human-centric
analysis. In this paper, we propose a segment-based parsing pipeline that
explores human pose information, i.e. the joint location of a human model,
which improves the part proposal, accelerates the inference and regularizes the
parsing process at the same time. Specifically, we first generate part segment
proposals with respect to human joints predicted by a deep model, then part-
specific ranking models are trained for segment selection using both pose-based
features and deep-learned part potential features. Finally, the best ensemble
of the proposed part segments are inferred though an And-Or Graph.
We evaluate our approach on the popular Penn-Fudan pedestrian parsing
dataset, and demonstrate the effectiveness of using the pose information for
each stage of the parsing pipeline. Finally, we show that our approach yields
superior part segmentation accuracy comparing to the state-of-the-art methods.
| [
{
"version": "v1",
"created": "Mon, 17 Aug 2015 00:05:38 GMT"
},
{
"version": "v2",
"created": "Wed, 25 Nov 2015 02:07:11 GMT"
}
] | 2015-11-26T00:00:00 | [
[
"Xia",
"Fangting",
""
],
[
"Zhu",
"Jun",
""
],
[
"Wang",
"Peng",
""
],
[
"Yuille",
"Alan",
""
]
] | TITLE: Pose-Guided Human Parsing with Deep Learned Features
ABSTRACT: Parsing human body into semantic regions is crucial to human-centric
analysis. In this paper, we propose a segment-based parsing pipeline that
explores human pose information, i.e. the joint location of a human model,
which improves the part proposal, accelerates the inference and regularizes the
parsing process at the same time. Specifically, we first generate part segment
proposals with respect to human joints predicted by a deep model, then part-
specific ranking models are trained for segment selection using both pose-based
features and deep-learned part potential features. Finally, the best ensemble
of the proposed part segments are inferred though an And-Or Graph.
We evaluate our approach on the popular Penn-Fudan pedestrian parsing
dataset, and demonstrate the effectiveness of using the pose information for
each stage of the parsing pipeline. Finally, we show that our approach yields
superior part segmentation accuracy comparing to the state-of-the-art methods.
| no_new_dataset | 0.952131 |
1511.05121 | Rahul Gopal Krishnan | Rahul G. Krishnan, Uri Shalit, David Sontag | Deep Kalman Filters | 17 pages, 14 figures: Fixed typo in Fig. 1(b) and added reference | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Kalman Filters are one of the most influential models of time-varying
phenomena. They admit an intuitive probabilistic interpretation, have a simple
functional form, and enjoy widespread adoption in a variety of disciplines.
Motivated by recent variational methods for learning deep generative models, we
introduce a unified algorithm to efficiently learn a broad spectrum of Kalman
filters. Of particular interest is the use of temporal generative models for
counterfactual inference. We investigate the efficacy of such models for
counterfactual inference, and to that end we introduce the "Healing MNIST"
dataset where long-term structure, noise and actions are applied to sequences
of digits. We show the efficacy of our method for modeling this dataset. We
further show how our model can be used for counterfactual inference for
patients, based on electronic health record data of 8,000 patients over 4.5
years.
| [
{
"version": "v1",
"created": "Mon, 16 Nov 2015 20:46:38 GMT"
},
{
"version": "v2",
"created": "Wed, 25 Nov 2015 20:47:00 GMT"
}
] | 2015-11-26T00:00:00 | [
[
"Krishnan",
"Rahul G.",
""
],
[
"Shalit",
"Uri",
""
],
[
"Sontag",
"David",
""
]
] | TITLE: Deep Kalman Filters
ABSTRACT: Kalman Filters are one of the most influential models of time-varying
phenomena. They admit an intuitive probabilistic interpretation, have a simple
functional form, and enjoy widespread adoption in a variety of disciplines.
Motivated by recent variational methods for learning deep generative models, we
introduce a unified algorithm to efficiently learn a broad spectrum of Kalman
filters. Of particular interest is the use of temporal generative models for
counterfactual inference. We investigate the efficacy of such models for
counterfactual inference, and to that end we introduce the "Healing MNIST"
dataset where long-term structure, noise and actions are applied to sequences
of digits. We show the efficacy of our method for modeling this dataset. We
further show how our model can be used for counterfactual inference for
patients, based on electronic health record data of 8,000 patients over 4.5
years.
| new_dataset | 0.96157 |
1511.07917 | Anton Osokin | Tuan-Hung Vu, Anton Osokin, Ivan Laptev | Context-aware CNNs for person head detection | To appear in International Conference on Computer Vision (ICCV), 2015 | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Person detection is a key problem for many computer vision tasks. While face
detection has reached maturity, detecting people under a full variation of
camera view-points, human poses, lighting conditions and occlusions is still a
difficult challenge. In this work we focus on detecting human heads in natural
scenes. Starting from the recent local R-CNN object detector, we extend it with
two types of contextual cues. First, we leverage person-scene relations and
propose a Global CNN model trained to predict positions and scales of heads
directly from the full image. Second, we explicitly model pairwise relations
among objects and train a Pairwise CNN model using a structured-output
surrogate loss. The Local, Global and Pairwise models are combined into a joint
CNN framework. To train and test our full model, we introduce a large dataset
composed of 369,846 human heads annotated in 224,740 movie frames. We evaluate
our method and demonstrate improvements of person head detection against
several recent baselines in three datasets. We also show improvements of the
detection speed provided by our model.
| [
{
"version": "v1",
"created": "Tue, 24 Nov 2015 23:23:18 GMT"
}
] | 2015-11-26T00:00:00 | [
[
"Vu",
"Tuan-Hung",
""
],
[
"Osokin",
"Anton",
""
],
[
"Laptev",
"Ivan",
""
]
] | TITLE: Context-aware CNNs for person head detection
ABSTRACT: Person detection is a key problem for many computer vision tasks. While face
detection has reached maturity, detecting people under a full variation of
camera view-points, human poses, lighting conditions and occlusions is still a
difficult challenge. In this work we focus on detecting human heads in natural
scenes. Starting from the recent local R-CNN object detector, we extend it with
two types of contextual cues. First, we leverage person-scene relations and
propose a Global CNN model trained to predict positions and scales of heads
directly from the full image. Second, we explicitly model pairwise relations
among objects and train a Pairwise CNN model using a structured-output
surrogate loss. The Local, Global and Pairwise models are combined into a joint
CNN framework. To train and test our full model, we introduce a large dataset
composed of 369,846 human heads annotated in 224,740 movie frames. We evaluate
our method and demonstrate improvements of person head detection against
several recent baselines in three datasets. We also show improvements of the
detection speed provided by our model.
| new_dataset | 0.957912 |
1511.07951 | Vittal Premachandran | Vittal Premachandran, Boyan Bonev, Alan L. Yuille | PASCAL Boundaries: A Class-Agnostic Semantic Boundary Dataset | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we address the boundary detection task motivated by the
ambiguities in current definition of edge detection. To this end, we generate a
large database consisting of more than 10k images (which is 20x bigger than
existing edge detection databases) along with ground truth boundaries between
459 semantic classes including both foreground objects and different types of
background, and call it the PASCAL Boundaries dataset, which will be released
to the community. In addition, we propose a novel deep network-based
multi-scale semantic boundary detector and name it Multi-scale Deep Semantic
Boundary Detector (M-DSBD). We provide baselines using models that were trained
on edge detection and show that they transfer reasonably to the task of
boundary detection. Finally, we point to various important research problems
that this dataset can be used for.
| [
{
"version": "v1",
"created": "Wed, 25 Nov 2015 05:12:38 GMT"
}
] | 2015-11-26T00:00:00 | [
[
"Premachandran",
"Vittal",
""
],
[
"Bonev",
"Boyan",
""
],
[
"Yuille",
"Alan L.",
""
]
] | TITLE: PASCAL Boundaries: A Class-Agnostic Semantic Boundary Dataset
ABSTRACT: In this paper, we address the boundary detection task motivated by the
ambiguities in current definition of edge detection. To this end, we generate a
large database consisting of more than 10k images (which is 20x bigger than
existing edge detection databases) along with ground truth boundaries between
459 semantic classes including both foreground objects and different types of
background, and call it the PASCAL Boundaries dataset, which will be released
to the community. In addition, we propose a novel deep network-based
multi-scale semantic boundary detector and name it Multi-scale Deep Semantic
Boundary Detector (M-DSBD). We provide baselines using models that were trained
on edge detection and show that they transfer reasonably to the task of
boundary detection. Finally, we point to various important research problems
that this dataset can be used for.
| new_dataset | 0.955026 |
1511.08032 | Christos Tzelepis | Christos Tzelepis, Damianos Galanopoulos, Vasileios Mezaris, Ioannis
Patras | Learning to detect video events from zero or very few video examples | Image and Vision Computing Journal, Elsevier, 2015, accepted for
publication | Image and Vision Computing Journal, Elsevier, 2015 | 10.1016/j.imavis.2015.09.005 | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work we deal with the problem of high-level event detection in video.
Specifically, we study the challenging problems of i) learning to detect video
events from solely a textual description of the event, without using any
positive video examples, and ii) additionally exploiting very few positive
training samples together with a small number of ``related'' videos. For
learning only from an event's textual description, we first identify a general
learning framework and then study the impact of different design choices for
various stages of this framework. For additionally learning from example
videos, when true positive training samples are scarce, we employ an extension
of the Support Vector Machine that allows us to exploit ``related'' event
videos by automatically introducing different weights for subsets of the videos
in the overall training set. Experimental evaluations performed on the
large-scale TRECVID MED 2014 video dataset provide insight on the effectiveness
of the proposed methods.
| [
{
"version": "v1",
"created": "Wed, 25 Nov 2015 12:17:50 GMT"
}
] | 2015-11-26T00:00:00 | [
[
"Tzelepis",
"Christos",
""
],
[
"Galanopoulos",
"Damianos",
""
],
[
"Mezaris",
"Vasileios",
""
],
[
"Patras",
"Ioannis",
""
]
] | TITLE: Learning to detect video events from zero or very few video examples
ABSTRACT: In this work we deal with the problem of high-level event detection in video.
Specifically, we study the challenging problems of i) learning to detect video
events from solely a textual description of the event, without using any
positive video examples, and ii) additionally exploiting very few positive
training samples together with a small number of ``related'' videos. For
learning only from an event's textual description, we first identify a general
learning framework and then study the impact of different design choices for
various stages of this framework. For additionally learning from example
videos, when true positive training samples are scarce, we employ an extension
of the Support Vector Machine that allows us to exploit ``related'' event
videos by automatically introducing different weights for subsets of the videos
in the overall training set. Experimental evaluations performed on the
large-scale TRECVID MED 2014 video dataset provide insight on the effectiveness
of the proposed methods.
| no_new_dataset | 0.950595 |
1511.08177 | Saurabh Gupta | Saurabh Gupta, Bharath Hariharan, Jitendra Malik | Exploring Person Context and Local Scene Context for Object Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we explore two ways of using context for object detection. The
first model focusses on people and the objects they commonly interact with,
such as fashion and sports accessories. The second model considers more general
object detection and uses the spatial relationships between objects and between
objects and scenes. Our models are able to capture precise spatial
relationships between the context and the object of interest, and make
effective use of the appearance of the contextual region. On the newly released
COCO dataset, our models provide relative improvements of up to 5% over
CNN-based state-of-the-art detectors, with the gains concentrated on hard cases
such as small objects (10% relative improvement).
| [
{
"version": "v1",
"created": "Wed, 25 Nov 2015 19:45:03 GMT"
}
] | 2015-11-26T00:00:00 | [
[
"Gupta",
"Saurabh",
""
],
[
"Hariharan",
"Bharath",
""
],
[
"Malik",
"Jitendra",
""
]
] | TITLE: Exploring Person Context and Local Scene Context for Object Detection
ABSTRACT: In this paper we explore two ways of using context for object detection. The
first model focusses on people and the objects they commonly interact with,
such as fashion and sports accessories. The second model considers more general
object detection and uses the spatial relationships between objects and between
objects and scenes. Our models are able to capture precise spatial
relationships between the context and the object of interest, and make
effective use of the appearance of the contextual region. On the newly released
COCO dataset, our models provide relative improvements of up to 5% over
CNN-based state-of-the-art detectors, with the gains concentrated on hard cases
such as small objects (10% relative improvement).
| new_dataset | 0.951953 |
1412.1526 | Xianjie Chen | Xianjie Chen, Alan Yuille | Parsing Occluded People by Flexible Compositions | CVPR 15 Camera Ready | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents an approach to parsing humans when there is significant
occlusion. We model humans using a graphical model which has a tree structure
building on recent work [32, 6] and exploit the connectivity prior that, even
in presence of occlusion, the visible nodes form a connected subtree of the
graphical model. We call each connected subtree a flexible composition of
object parts. This involves a novel method for learning occlusion cues. During
inference we need to search over a mixture of different flexible models. By
exploiting part sharing, we show that this inference can be done extremely
efficiently requiring only twice as many computations as searching for the
entire object (i.e., not modeling occlusion). We evaluate our model on the
standard benchmarked "We Are Family" Stickmen dataset and obtain significant
performance improvements over the best alternative algorithms.
| [
{
"version": "v1",
"created": "Thu, 4 Dec 2014 00:45:14 GMT"
},
{
"version": "v2",
"created": "Tue, 24 Nov 2015 07:57:19 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Chen",
"Xianjie",
""
],
[
"Yuille",
"Alan",
""
]
] | TITLE: Parsing Occluded People by Flexible Compositions
ABSTRACT: This paper presents an approach to parsing humans when there is significant
occlusion. We model humans using a graphical model which has a tree structure
building on recent work [32, 6] and exploit the connectivity prior that, even
in presence of occlusion, the visible nodes form a connected subtree of the
graphical model. We call each connected subtree a flexible composition of
object parts. This involves a novel method for learning occlusion cues. During
inference we need to search over a mixture of different flexible models. By
exploiting part sharing, we show that this inference can be done extremely
efficiently requiring only twice as many computations as searching for the
entire object (i.e., not modeling occlusion). We evaluate our model on the
standard benchmarked "We Are Family" Stickmen dataset and obtain significant
performance improvements over the best alternative algorithms.
| no_new_dataset | 0.947575 |
1503.08895 | Sainbayar Sukhbaatar | Sainbayar Sukhbaatar, Arthur Szlam, Jason Weston and Rob Fergus | End-To-End Memory Networks | Accepted to NIPS 2015 | null | null | null | cs.NE cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a neural network with a recurrent attention model over a
possibly large external memory. The architecture is a form of Memory Network
(Weston et al., 2015) but unlike the model in that work, it is trained
end-to-end, and hence requires significantly less supervision during training,
making it more generally applicable in realistic settings. It can also be seen
as an extension of RNNsearch to the case where multiple computational steps
(hops) are performed per output symbol. The flexibility of the model allows us
to apply it to tasks as diverse as (synthetic) question answering and to
language modeling. For the former our approach is competitive with Memory
Networks, but with less supervision. For the latter, on the Penn TreeBank and
Text8 datasets our approach demonstrates comparable performance to RNNs and
LSTMs. In both cases we show that the key concept of multiple computational
hops yields improved results.
| [
{
"version": "v1",
"created": "Tue, 31 Mar 2015 03:05:37 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Apr 2015 02:23:20 GMT"
},
{
"version": "v3",
"created": "Sun, 12 Apr 2015 04:19:33 GMT"
},
{
"version": "v4",
"created": "Mon, 8 Jun 2015 21:42:20 GMT"
},
{
"version": "v5",
"created": "Tue, 24 Nov 2015 19:41:57 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Sukhbaatar",
"Sainbayar",
""
],
[
"Szlam",
"Arthur",
""
],
[
"Weston",
"Jason",
""
],
[
"Fergus",
"Rob",
""
]
] | TITLE: End-To-End Memory Networks
ABSTRACT: We introduce a neural network with a recurrent attention model over a
possibly large external memory. The architecture is a form of Memory Network
(Weston et al., 2015) but unlike the model in that work, it is trained
end-to-end, and hence requires significantly less supervision during training,
making it more generally applicable in realistic settings. It can also be seen
as an extension of RNNsearch to the case where multiple computational steps
(hops) are performed per output symbol. The flexibility of the model allows us
to apply it to tasks as diverse as (synthetic) question answering and to
language modeling. For the former our approach is competitive with Memory
Networks, but with less supervision. For the latter, on the Penn TreeBank and
Text8 datasets our approach demonstrates comparable performance to RNNs and
LSTMs. In both cases we show that the key concept of multiple computational
hops yields improved results.
| no_new_dataset | 0.953708 |
1505.02438 | Stavros Tsogkas | S. Tsogkas, I. Kokkinos, G. Papandreou, A. Vedaldi | Deep Learning for Semantic Part Segmentation with High-Level Guidance | 11 pages (including references), 3 figures, 2 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work we address the task of segmenting an object into its parts, or
semantic part segmentation. We start by adapting a state-of-the-art semantic
segmentation system to this task, and show that a combination of a
fully-convolutional Deep CNN system coupled with Dense CRF labelling provides
excellent results for a broad range of object categories. Still, this approach
remains agnostic to high-level constraints between object parts. We introduce
such prior information by means of the Restricted Boltzmann Machine, adapted to
our task and train our model in an discriminative fashion, as a hidden CRF,
demonstrating that prior information can yield additional improvements. We also
investigate the performance of our approach ``in the wild'', without
information concerning the objects' bounding boxes, using an object detector to
guide a multi-scale segmentation scheme. We evaluate the performance of our
approach on the Penn-Fudan and LFW datasets for the tasks of pedestrian parsing
and face labelling respectively. We show superior performance with respect to
competitive methods that have been extensively engineered on these benchmarks,
as well as realistic qualitative results on part segmentation, even for
occluded or deformable objects. We also provide quantitative and extensive
qualitative results on three classes from the PASCAL Parts dataset. Finally, we
show that our multi-scale segmentation scheme can boost accuracy, recovering
segmentations for finer parts.
| [
{
"version": "v1",
"created": "Sun, 10 May 2015 21:12:31 GMT"
},
{
"version": "v2",
"created": "Tue, 24 Nov 2015 14:22:43 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Tsogkas",
"S.",
""
],
[
"Kokkinos",
"I.",
""
],
[
"Papandreou",
"G.",
""
],
[
"Vedaldi",
"A.",
""
]
] | TITLE: Deep Learning for Semantic Part Segmentation with High-Level Guidance
ABSTRACT: In this work we address the task of segmenting an object into its parts, or
semantic part segmentation. We start by adapting a state-of-the-art semantic
segmentation system to this task, and show that a combination of a
fully-convolutional Deep CNN system coupled with Dense CRF labelling provides
excellent results for a broad range of object categories. Still, this approach
remains agnostic to high-level constraints between object parts. We introduce
such prior information by means of the Restricted Boltzmann Machine, adapted to
our task and train our model in an discriminative fashion, as a hidden CRF,
demonstrating that prior information can yield additional improvements. We also
investigate the performance of our approach ``in the wild'', without
information concerning the objects' bounding boxes, using an object detector to
guide a multi-scale segmentation scheme. We evaluate the performance of our
approach on the Penn-Fudan and LFW datasets for the tasks of pedestrian parsing
and face labelling respectively. We show superior performance with respect to
competitive methods that have been extensively engineered on these benchmarks,
as well as realistic qualitative results on part segmentation, even for
occluded or deformable objects. We also provide quantitative and extensive
qualitative results on three classes from the PASCAL Parts dataset. Finally, we
show that our multi-scale segmentation scheme can boost accuracy, recovering
segmentations for finer parts.
| no_new_dataset | 0.946448 |
1511.03371 | Jack Hessel | Jack Hessel, Alexandra Schofield, Lillian Lee, David Mimno | What do Vegans do in their Spare Time? Latent Interest Detection in
Multi-Community Networks | NIPS 2015 Network Workshop | null | null | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most social network analysis works at the level of interactions between
users. But the vast growth in size and complexity of social networks enables us
to examine interactions at larger scale. In this work we use a dataset of 76M
submissions to the social network Reddit, which is organized into distinct
sub-communities called subreddits. We measure the similarity between entire
subreddits both in terms of user similarity and topical similarity. Our goal is
to find community pairs with similar userbases, but dissimilar content; we
refer to this type of relationship as a "latent interest." Detection of latent
interests not only provides a perspective on individual users as they shift
between roles (student, sports fan, political activist) but also gives insight
into the dynamics of Reddit as a whole. Latent interest detection also has
potential applications for recommendation systems and for researchers examining
community evolution.
| [
{
"version": "v1",
"created": "Wed, 11 Nov 2015 03:07:00 GMT"
},
{
"version": "v2",
"created": "Tue, 24 Nov 2015 05:11:11 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Hessel",
"Jack",
""
],
[
"Schofield",
"Alexandra",
""
],
[
"Lee",
"Lillian",
""
],
[
"Mimno",
"David",
""
]
] | TITLE: What do Vegans do in their Spare Time? Latent Interest Detection in
Multi-Community Networks
ABSTRACT: Most social network analysis works at the level of interactions between
users. But the vast growth in size and complexity of social networks enables us
to examine interactions at larger scale. In this work we use a dataset of 76M
submissions to the social network Reddit, which is organized into distinct
sub-communities called subreddits. We measure the similarity between entire
subreddits both in terms of user similarity and topical similarity. Our goal is
to find community pairs with similar userbases, but dissimilar content; we
refer to this type of relationship as a "latent interest." Detection of latent
interests not only provides a perspective on individual users as they shift
between roles (student, sports fan, political activist) but also gives insight
into the dynamics of Reddit as a whole. Latent interest detection also has
potential applications for recommendation systems and for researchers examining
community evolution.
| no_new_dataset | 0.903038 |
1511.06891 | Kian Hsiang Low | Yehong Zhang, Trong Nghia Hoang, Kian Hsiang Low, Mohan Kankanhalli | Near-Optimal Active Learning of Multi-Output Gaussian Processes | 30th AAAI Conference on Artificial Intelligence (AAAI 2016), Extended
version with proofs, 13 pages | null | null | null | stat.ML cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper addresses the problem of active learning of a multi-output
Gaussian process (MOGP) model representing multiple types of coexisting
correlated environmental phenomena. In contrast to existing works, our active
learning problem involves selecting not just the most informative sampling
locations to be observed but also the types of measurements at each selected
location for minimizing the predictive uncertainty (i.e., posterior joint
entropy) of a target phenomenon of interest given a sampling budget.
Unfortunately, such an entropy criterion scales poorly in the numbers of
candidate sampling locations and selected observations when optimized. To
resolve this issue, we first exploit a structure common to sparse MOGP models
for deriving a novel active learning criterion. Then, we exploit a relaxed form
of submodularity property of our new criterion for devising a polynomial-time
approximation algorithm that guarantees a constant-factor approximation of that
achieved by the optimal set of selected observations. Empirical evaluation on
real-world datasets shows that our proposed approach outperforms existing
algorithms for active learning of MOGP and single-output GP models.
| [
{
"version": "v1",
"created": "Sat, 21 Nov 2015 15:08:53 GMT"
},
{
"version": "v2",
"created": "Tue, 24 Nov 2015 08:45:36 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Zhang",
"Yehong",
""
],
[
"Hoang",
"Trong Nghia",
""
],
[
"Low",
"Kian Hsiang",
""
],
[
"Kankanhalli",
"Mohan",
""
]
] | TITLE: Near-Optimal Active Learning of Multi-Output Gaussian Processes
ABSTRACT: This paper addresses the problem of active learning of a multi-output
Gaussian process (MOGP) model representing multiple types of coexisting
correlated environmental phenomena. In contrast to existing works, our active
learning problem involves selecting not just the most informative sampling
locations to be observed but also the types of measurements at each selected
location for minimizing the predictive uncertainty (i.e., posterior joint
entropy) of a target phenomenon of interest given a sampling budget.
Unfortunately, such an entropy criterion scales poorly in the numbers of
candidate sampling locations and selected observations when optimized. To
resolve this issue, we first exploit a structure common to sparse MOGP models
for deriving a novel active learning criterion. Then, we exploit a relaxed form
of submodularity property of our new criterion for devising a polynomial-time
approximation algorithm that guarantees a constant-factor approximation of that
achieved by the optimal set of selected observations. Empirical evaluation on
real-world datasets shows that our proposed approach outperforms existing
algorithms for active learning of MOGP and single-output GP models.
| no_new_dataset | 0.945951 |
1511.07528 | Nicolas Papernot | Nicolas Papernot and Patrick McDaniel and Somesh Jha and Matt
Fredrikson and Z. Berkay Celik and Ananthram Swami | The Limitations of Deep Learning in Adversarial Settings | Accepted to the 1st IEEE European Symposium on Security & Privacy,
IEEE 2016. Saarbrucken, Germany | null | null | null | cs.CR cs.LG cs.NE stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep learning takes advantage of large datasets and computationally efficient
training algorithms to outperform other approaches at various machine learning
tasks. However, imperfections in the training phase of deep neural networks
make them vulnerable to adversarial samples: inputs crafted by adversaries with
the intent of causing deep neural networks to misclassify. In this work, we
formalize the space of adversaries against deep neural networks (DNNs) and
introduce a novel class of algorithms to craft adversarial samples based on a
precise understanding of the mapping between inputs and outputs of DNNs. In an
application to computer vision, we show that our algorithms can reliably
produce samples correctly classified by human subjects but misclassified in
specific targets by a DNN with a 97% adversarial success rate while only
modifying on average 4.02% of the input features per sample. We then evaluate
the vulnerability of different sample classes to adversarial perturbations by
defining a hardness measure. Finally, we describe preliminary work outlining
defenses against adversarial samples by defining a predictive measure of
distance between a benign input and a target classification.
| [
{
"version": "v1",
"created": "Tue, 24 Nov 2015 01:07:08 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Papernot",
"Nicolas",
""
],
[
"McDaniel",
"Patrick",
""
],
[
"Jha",
"Somesh",
""
],
[
"Fredrikson",
"Matt",
""
],
[
"Celik",
"Z. Berkay",
""
],
[
"Swami",
"Ananthram",
""
]
] | TITLE: The Limitations of Deep Learning in Adversarial Settings
ABSTRACT: Deep learning takes advantage of large datasets and computationally efficient
training algorithms to outperform other approaches at various machine learning
tasks. However, imperfections in the training phase of deep neural networks
make them vulnerable to adversarial samples: inputs crafted by adversaries with
the intent of causing deep neural networks to misclassify. In this work, we
formalize the space of adversaries against deep neural networks (DNNs) and
introduce a novel class of algorithms to craft adversarial samples based on a
precise understanding of the mapping between inputs and outputs of DNNs. In an
application to computer vision, we show that our algorithms can reliably
produce samples correctly classified by human subjects but misclassified in
specific targets by a DNN with a 97% adversarial success rate while only
modifying on average 4.02% of the input features per sample. We then evaluate
the vulnerability of different sample classes to adversarial perturbations by
defining a hardness measure. Finally, we describe preliminary work outlining
defenses against adversarial samples by defining a predictive measure of
distance between a benign input and a target classification.
| no_new_dataset | 0.942135 |
1511.07571 | Justin Johnson | Justin Johnson and Andrej Karpathy and Li Fei-Fei | DenseCap: Fully Convolutional Localization Networks for Dense Captioning | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce the dense captioning task, which requires a computer vision
system to both localize and describe salient regions in images in natural
language. The dense captioning task generalizes object detection when the
descriptions consist of a single word, and Image Captioning when one predicted
region covers the full image. To address the localization and description task
jointly we propose a Fully Convolutional Localization Network (FCLN)
architecture that processes an image with a single, efficient forward pass,
requires no external regions proposals, and can be trained end-to-end with a
single round of optimization. The architecture is composed of a Convolutional
Network, a novel dense localization layer, and Recurrent Neural Network
language model that generates the label sequences. We evaluate our network on
the Visual Genome dataset, which comprises 94,000 images and 4,100,000
region-grounded captions. We observe both speed and accuracy improvements over
baselines based on current state of the art approaches in both generation and
retrieval settings.
| [
{
"version": "v1",
"created": "Tue, 24 Nov 2015 05:13:54 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Johnson",
"Justin",
""
],
[
"Karpathy",
"Andrej",
""
],
[
"Fei-Fei",
"Li",
""
]
] | TITLE: DenseCap: Fully Convolutional Localization Networks for Dense Captioning
ABSTRACT: We introduce the dense captioning task, which requires a computer vision
system to both localize and describe salient regions in images in natural
language. The dense captioning task generalizes object detection when the
descriptions consist of a single word, and Image Captioning when one predicted
region covers the full image. To address the localization and description task
jointly we propose a Fully Convolutional Localization Network (FCLN)
architecture that processes an image with a single, efficient forward pass,
requires no external regions proposals, and can be trained end-to-end with a
single round of optimization. The architecture is composed of a Convolutional
Network, a novel dense localization layer, and Recurrent Neural Network
language model that generates the label sequences. We evaluate our network on
the Visual Genome dataset, which comprises 94,000 images and 4,100,000
region-grounded captions. We observe both speed and accuracy improvements over
baselines based on current state of the art approaches in both generation and
retrieval settings.
| no_new_dataset | 0.948442 |
1511.07732 | Thiago Santini | Thiago Santini, Wolfgang Fuhl, Thomas K\"ubler, and Enkelejda Kasneci | Bayesian Identification of Fixations, Saccades, and Smooth Pursuits | 8 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Smooth pursuit eye movements provide meaningful insights and information on
subject's behavior and health and may, in particular situations, disturb the
performance of typical fixation/saccade classification algorithms. Thus, an
automatic and efficient algorithm to identify these eye movements is paramount
for eye-tracking research involving dynamic stimuli. In this paper, we propose
the Bayesian Decision Theory Identification (I-BDT) algorithm, a novel
algorithm for ternary classification of eye movements that is able to reliably
separate fixations, saccades, and smooth pursuits in an online fashion, even
for low-resolution eye trackers. The proposed algorithm is evaluated on four
datasets with distinct mixtures of eye movements, including fixations,
saccades, as well as straight and circular smooth pursuits; data was collected
with a sample rate of 30 Hz from six subjects, totaling 24 evaluation datasets.
The algorithm exhibits high and consistent performance across all datasets and
movements relative to a manual annotation by a domain expert (recall: \mu =
91.42%, \sigma = 9.52%; precision: \mu = 95.60%, \sigma = 5.29%; specificity
\mu = 95.41%, \sigma = 7.02%) and displays a significant improvement when
compared to I-VDT, an state-of-the-art algorithm (recall: \mu = 87.67%, \sigma
= 14.73%; precision: \mu = 89.57%, \sigma = 8.05%; specificity \mu = 92.10%,
\sigma = 11.21%). For algorithm implementation and annotated datasets, please
contact the first author.
| [
{
"version": "v1",
"created": "Tue, 24 Nov 2015 14:40:05 GMT"
}
] | 2015-11-25T00:00:00 | [
[
"Santini",
"Thiago",
""
],
[
"Fuhl",
"Wolfgang",
""
],
[
"Kübler",
"Thomas",
""
],
[
"Kasneci",
"Enkelejda",
""
]
] | TITLE: Bayesian Identification of Fixations, Saccades, and Smooth Pursuits
ABSTRACT: Smooth pursuit eye movements provide meaningful insights and information on
subject's behavior and health and may, in particular situations, disturb the
performance of typical fixation/saccade classification algorithms. Thus, an
automatic and efficient algorithm to identify these eye movements is paramount
for eye-tracking research involving dynamic stimuli. In this paper, we propose
the Bayesian Decision Theory Identification (I-BDT) algorithm, a novel
algorithm for ternary classification of eye movements that is able to reliably
separate fixations, saccades, and smooth pursuits in an online fashion, even
for low-resolution eye trackers. The proposed algorithm is evaluated on four
datasets with distinct mixtures of eye movements, including fixations,
saccades, as well as straight and circular smooth pursuits; data was collected
with a sample rate of 30 Hz from six subjects, totaling 24 evaluation datasets.
The algorithm exhibits high and consistent performance across all datasets and
movements relative to a manual annotation by a domain expert (recall: \mu =
91.42%, \sigma = 9.52%; precision: \mu = 95.60%, \sigma = 5.29%; specificity
\mu = 95.41%, \sigma = 7.02%) and displays a significant improvement when
compared to I-VDT, an state-of-the-art algorithm (recall: \mu = 87.67%, \sigma
= 14.73%; precision: \mu = 89.57%, \sigma = 8.05%; specificity \mu = 92.10%,
\sigma = 11.21%). For algorithm implementation and annotated datasets, please
contact the first author.
| no_new_dataset | 0.954393 |
1501.04709 | Boleslaw Szymanski | Chris Gaiteri, Mingming Chen, Boleslaw Szymanski, Konstantin Kuzmin,
Jierui Xie, Changkyu Lee, Timothy Blanche, Elias Chaibub Neto, Su-Chun Huang,
Thomas Grabowski, Tara Madhyastha, Vitalina Komashko | Identifying robust communities and multi-community nodes by combining
top-down and bottom-up approaches to clustering | null | Scientific Reports 5, Article number: 16361 (2015) | 10.1038/srep16361 | null | cs.CE cs.SI physics.bio-ph q-bio.MN | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Biological functions are carried out by groups of interacting molecules,
cells or tissues, known as communities. Membership in these communities may
overlap when biological components are involved in multiple functions. However,
traditional clustering methods detect non-overlapping communities. These
detected communities may also be unstable and difficult to replicate, because
traditional methods are sensitive to noise and parameter settings. These
aspects of traditional clustering methods limit our ability to detect
biological communities, and therefore our ability to understand biological
functions.
To address these limitations and detect robust overlapping biological
communities, we propose an unorthodox clustering method called SpeakEasy which
identifies communities using top-down and bottom-up approaches simultaneously.
Specifically, nodes join communities based on their local connections, as well
as global information about the network structure. This method can quantify the
stability of each community, automatically identify the number of communities,
and quickly cluster networks with hundreds of thousands of nodes.
SpeakEasy shows top performance on synthetic clustering benchmarks and
accurately identifies meaningful biological communities in a range of datasets,
including: gene microarrays, protein interactions, sorted cell populations,
electrophysiology and fMRI brain imaging.
| [
{
"version": "v1",
"created": "Tue, 20 Jan 2015 04:15:17 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Feb 2015 00:30:22 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Gaiteri",
"Chris",
""
],
[
"Chen",
"Mingming",
""
],
[
"Szymanski",
"Boleslaw",
""
],
[
"Kuzmin",
"Konstantin",
""
],
[
"Xie",
"Jierui",
""
],
[
"Lee",
"Changkyu",
""
],
[
"Blanche",
"Timothy",
""
],
[
"Neto",
"Elias Chaibub",
""
],
[
"Huang",
"Su-Chun",
""
],
[
"Grabowski",
"Thomas",
""
],
[
"Madhyastha",
"Tara",
""
],
[
"Komashko",
"Vitalina",
""
]
] | TITLE: Identifying robust communities and multi-community nodes by combining
top-down and bottom-up approaches to clustering
ABSTRACT: Biological functions are carried out by groups of interacting molecules,
cells or tissues, known as communities. Membership in these communities may
overlap when biological components are involved in multiple functions. However,
traditional clustering methods detect non-overlapping communities. These
detected communities may also be unstable and difficult to replicate, because
traditional methods are sensitive to noise and parameter settings. These
aspects of traditional clustering methods limit our ability to detect
biological communities, and therefore our ability to understand biological
functions.
To address these limitations and detect robust overlapping biological
communities, we propose an unorthodox clustering method called SpeakEasy which
identifies communities using top-down and bottom-up approaches simultaneously.
Specifically, nodes join communities based on their local connections, as well
as global information about the network structure. This method can quantify the
stability of each community, automatically identify the number of communities,
and quickly cluster networks with hundreds of thousands of nodes.
SpeakEasy shows top performance on synthetic clustering benchmarks and
accurately identifies meaningful biological communities in a range of datasets,
including: gene microarrays, protein interactions, sorted cell populations,
electrophysiology and fMRI brain imaging.
| no_new_dataset | 0.945349 |
1505.05836 | Aroma Mahendru | Neelima Chavali, Harsh Agrawal, Aroma Mahendru, Dhruv Batra | Object-Proposal Evaluation Protocol is 'Gameable' | 15 pages, 11 figures, 4 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object proposals have quickly become the de-facto pre-processing step in a
number of vision pipelines (for object detection, object discovery, and other
tasks). Their performance is usually evaluated on partially annotated datasets.
In this paper, we argue that the choice of using a partially annotated dataset
for evaluation of object proposals is problematic -- as we demonstrate via a
thought experiment, the evaluation protocol is 'gameable', in the sense that
progress under this protocol does not necessarily correspond to a "better"
category independent object proposal algorithm.
To alleviate this problem, we: (1) Introduce a nearly-fully annotated version
of PASCAL VOC dataset, which serves as a test-bed to check if object proposal
techniques are overfitting to a particular list of categories. (2) Perform an
exhaustive evaluation of object proposal methods on our introduced nearly-fully
annotated PASCAL dataset and perform cross-dataset generalization experiments;
and (3) Introduce a diagnostic experiment to detect the bias capacity in an
object proposal algorithm. This tool circumvents the need to collect a densely
annotated dataset, which can be expensive and cumbersome to collect. Finally,
we plan to release an easy-to-use toolbox which combines various publicly
available implementations of object proposal algorithms which standardizes the
proposal generation and evaluation so that new methods can be added and
evaluated on different datasets. We hope that the results presented in the
paper will motivate the community to test the category independence of various
object proposal methods by carefully choosing the evaluation protocol.
| [
{
"version": "v1",
"created": "Thu, 21 May 2015 18:53:45 GMT"
},
{
"version": "v2",
"created": "Fri, 22 May 2015 04:23:57 GMT"
},
{
"version": "v3",
"created": "Mon, 25 May 2015 05:14:53 GMT"
},
{
"version": "v4",
"created": "Mon, 23 Nov 2015 07:16:16 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Chavali",
"Neelima",
""
],
[
"Agrawal",
"Harsh",
""
],
[
"Mahendru",
"Aroma",
""
],
[
"Batra",
"Dhruv",
""
]
] | TITLE: Object-Proposal Evaluation Protocol is 'Gameable'
ABSTRACT: Object proposals have quickly become the de-facto pre-processing step in a
number of vision pipelines (for object detection, object discovery, and other
tasks). Their performance is usually evaluated on partially annotated datasets.
In this paper, we argue that the choice of using a partially annotated dataset
for evaluation of object proposals is problematic -- as we demonstrate via a
thought experiment, the evaluation protocol is 'gameable', in the sense that
progress under this protocol does not necessarily correspond to a "better"
category independent object proposal algorithm.
To alleviate this problem, we: (1) Introduce a nearly-fully annotated version
of PASCAL VOC dataset, which serves as a test-bed to check if object proposal
techniques are overfitting to a particular list of categories. (2) Perform an
exhaustive evaluation of object proposal methods on our introduced nearly-fully
annotated PASCAL dataset and perform cross-dataset generalization experiments;
and (3) Introduce a diagnostic experiment to detect the bias capacity in an
object proposal algorithm. This tool circumvents the need to collect a densely
annotated dataset, which can be expensive and cumbersome to collect. Finally,
we plan to release an easy-to-use toolbox which combines various publicly
available implementations of object proposal algorithms which standardizes the
proposal generation and evaluation so that new methods can be added and
evaluated on different datasets. We hope that the results presented in the
paper will motivate the community to test the category independence of various
object proposal methods by carefully choosing the evaluation protocol.
| new_dataset | 0.969699 |
1507.03060 | Hongkai Yu | Hongkai Yu, Youjie Zhou, Hui Qian, Min Xian, Yuewei Lin, Dazhou Guo,
Kang Zheng, Kareem Abdelfatah, Song Wang | LooseCut: Interactive Image Segmentation with Loosely Bounded Boxes | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One popular approach to interactively segment the foreground object of
interest from an image is to annotate a bounding box that covers the foreground
object. Then, a binary labeling is performed to achieve a refined segmentation.
One major issue of the existing algorithms for such interactive image
segmentation is their preference of an input bounding box that tightly encloses
the foreground object. This increases the annotation burden, and prevents these
algorithms from utilizing automatically detected bounding boxes. In this paper,
we develop a new LooseCut algorithm that can handle cases where the input
bounding box only loosely covers the foreground object. We propose a new Markov
Random Fields (MRF) model for segmentation with loosely bounded boxes,
including a global similarity constraint to better distinguish the foreground
and background, and an additional energy term to encourage consistent labeling
of similar-appearance pixels. This MRF model is then solved by an iterated
max-flow algorithm. In the experiments, we evaluate LooseCut in three
publicly-available image datasets, and compare its performance against several
state-of-the-art interactive image segmentation algorithms. We also show that
LooseCut can be used for enhancing the performance of unsupervised video
segmentation and image saliency detection.
| [
{
"version": "v1",
"created": "Sat, 11 Jul 2015 03:04:36 GMT"
},
{
"version": "v2",
"created": "Sun, 22 Nov 2015 03:54:17 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Yu",
"Hongkai",
""
],
[
"Zhou",
"Youjie",
""
],
[
"Qian",
"Hui",
""
],
[
"Xian",
"Min",
""
],
[
"Lin",
"Yuewei",
""
],
[
"Guo",
"Dazhou",
""
],
[
"Zheng",
"Kang",
""
],
[
"Abdelfatah",
"Kareem",
""
],
[
"Wang",
"Song",
""
]
] | TITLE: LooseCut: Interactive Image Segmentation with Loosely Bounded Boxes
ABSTRACT: One popular approach to interactively segment the foreground object of
interest from an image is to annotate a bounding box that covers the foreground
object. Then, a binary labeling is performed to achieve a refined segmentation.
One major issue of the existing algorithms for such interactive image
segmentation is their preference of an input bounding box that tightly encloses
the foreground object. This increases the annotation burden, and prevents these
algorithms from utilizing automatically detected bounding boxes. In this paper,
we develop a new LooseCut algorithm that can handle cases where the input
bounding box only loosely covers the foreground object. We propose a new Markov
Random Fields (MRF) model for segmentation with loosely bounded boxes,
including a global similarity constraint to better distinguish the foreground
and background, and an additional energy term to encourage consistent labeling
of similar-appearance pixels. This MRF model is then solved by an iterated
max-flow algorithm. In the experiments, we evaluate LooseCut in three
publicly-available image datasets, and compare its performance against several
state-of-the-art interactive image segmentation algorithms. We also show that
LooseCut can be used for enhancing the performance of unsupervised video
segmentation and image saliency detection.
| no_new_dataset | 0.949856 |
1508.01292 | Ilya Kalinowski | Ilya Kalinovskii, Vladimir Spitsyn | Compact Convolutional Neural Network Cascade for Face Detection | Demo video and test results available at
http://github.com/Bkmz21/FD-Evaluation | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The problem of faces detection in images or video streams is a classical
problem of computer vision. The multiple solutions of this problem have been
proposed, but the question of their optimality is still open. Many algorithms
achieve a high quality face detection, but at the cost of high computational
complexity. This restricts their application in the real-time systems. This
paper presents a new solution of the frontal face detection problem based on
compact convolutional neural networks cascade. The test results on FDDB dataset
show that it is competitive with state-of-the-art algorithms. This proposed
detector is implemented using three technologies: SSE/AVX/AVX2 instruction sets
for Intel CPUs, Nvidia CUDA, OpenCL. The detection speed of our approach
considerably exceeds all the existing CPU-based and GPU-based algorithms.
Because of high computational efficiency, our detector can processing 4K Ultra
HD video stream in real time (up to 27 fps) on mobile platforms (Intel Ivy
Bridge CPUs and Nvidia Kepler GPUs) in searching objects with the dimension
60x60 pixels or higher. At the same time its performance weakly dependent on
the background and number of objects in scene. This is achieved by the
asynchronous computation of stages in the cascade.
| [
{
"version": "v1",
"created": "Thu, 6 Aug 2015 07:01:55 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Aug 2015 09:10:27 GMT"
},
{
"version": "v3",
"created": "Mon, 23 Nov 2015 20:01:06 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Kalinovskii",
"Ilya",
""
],
[
"Spitsyn",
"Vladimir",
""
]
] | TITLE: Compact Convolutional Neural Network Cascade for Face Detection
ABSTRACT: The problem of faces detection in images or video streams is a classical
problem of computer vision. The multiple solutions of this problem have been
proposed, but the question of their optimality is still open. Many algorithms
achieve a high quality face detection, but at the cost of high computational
complexity. This restricts their application in the real-time systems. This
paper presents a new solution of the frontal face detection problem based on
compact convolutional neural networks cascade. The test results on FDDB dataset
show that it is competitive with state-of-the-art algorithms. This proposed
detector is implemented using three technologies: SSE/AVX/AVX2 instruction sets
for Intel CPUs, Nvidia CUDA, OpenCL. The detection speed of our approach
considerably exceeds all the existing CPU-based and GPU-based algorithms.
Because of high computational efficiency, our detector can processing 4K Ultra
HD video stream in real time (up to 27 fps) on mobile platforms (Intel Ivy
Bridge CPUs and Nvidia Kepler GPUs) in searching objects with the dimension
60x60 pixels or higher. At the same time its performance weakly dependent on
the background and number of objects in scene. This is achieved by the
asynchronous computation of stages in the cascade.
| no_new_dataset | 0.946399 |
1511.05296 | Jinghua Wang | Jinghua Wang, Abrar Abdul Nabi, Gang Wang, Chengde Wan, Tian-Tsong Ng | Towards Predicting the Likeability of Fashion Images | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a method for ranking fashion images to find the
ones which might be liked by more people. We collect two new datasets from
image sharing websites (Pinterest and Polyvore). We represent fashion images
based on attributes: semantic attributes and data-driven attributes. To learn
semantic attributes from limited training data, we use an algorithm on
multi-task convolutional neural networks to share visual knowledge among
different semantic attribute categories. To discover data-driven attributes
unsupervisedly, we propose an algorithm to simultaneously discover visual
clusters and learn fashion-specific feature representations. Given attributes
as representations, we propose to learn a ranking SPN (sum product networks) to
rank pairs of fashion images. The proposed ranking SPN can capture the
high-order correlations of the attributes. We show the effectiveness of our
method on our two newly collected datasets.
| [
{
"version": "v1",
"created": "Tue, 17 Nov 2015 07:31:36 GMT"
},
{
"version": "v2",
"created": "Mon, 23 Nov 2015 07:26:25 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Wang",
"Jinghua",
""
],
[
"Nabi",
"Abrar Abdul",
""
],
[
"Wang",
"Gang",
""
],
[
"Wan",
"Chengde",
""
],
[
"Ng",
"Tian-Tsong",
""
]
] | TITLE: Towards Predicting the Likeability of Fashion Images
ABSTRACT: In this paper, we propose a method for ranking fashion images to find the
ones which might be liked by more people. We collect two new datasets from
image sharing websites (Pinterest and Polyvore). We represent fashion images
based on attributes: semantic attributes and data-driven attributes. To learn
semantic attributes from limited training data, we use an algorithm on
multi-task convolutional neural networks to share visual knowledge among
different semantic attribute categories. To discover data-driven attributes
unsupervisedly, we propose an algorithm to simultaneously discover visual
clusters and learn fashion-specific feature representations. Given attributes
as representations, we propose to learn a ranking SPN (sum product networks) to
rank pairs of fashion images. The proposed ranking SPN can capture the
high-order correlations of the attributes. We show the effectiveness of our
method on our two newly collected datasets.
| new_dataset | 0.927495 |
1511.06833 | Thenmalar S | S. Thenmalar, J. Balaji, and T.V. Geetha | Semi-supervised Bootstrapping approach for Named Entity Recognition | 13 pages, 2 figures, 5 tables | null | null | null | cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The aim of Named Entity Recognition (NER) is to identify references of named
entities in unstructured documents, and to classify them into pre-defined
semantic categories. NER often aids from added background knowledge in the form
of gazetteers. However using such a collection does not deal with name variants
and cannot resolve ambiguities associated in identifying the entities in
context and associating them with predefined categories. We present a
semi-supervised NER approach that starts with identifying named entities with a
small set of training data. Using the identified named entities, the word and
the context features are used to define the pattern. This pattern of each named
entity category is used as a seed pattern to identify the named entities in the
test set. Pattern scoring and tuple value score enables the generation of the
new patterns to identify the named entity categories. We have evaluated the
proposed system for English language with the dataset of tagged (IEER) and
untagged (CoNLL 2003) named entity corpus and for Tamil language with the
documents from the FIRE corpus and yield an average f-measure of 75% for both
the languages.
| [
{
"version": "v1",
"created": "Sat, 21 Nov 2015 04:11:44 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Thenmalar",
"S.",
""
],
[
"Balaji",
"J.",
""
],
[
"Geetha",
"T. V.",
""
]
] | TITLE: Semi-supervised Bootstrapping approach for Named Entity Recognition
ABSTRACT: The aim of Named Entity Recognition (NER) is to identify references of named
entities in unstructured documents, and to classify them into pre-defined
semantic categories. NER often aids from added background knowledge in the form
of gazetteers. However using such a collection does not deal with name variants
and cannot resolve ambiguities associated in identifying the entities in
context and associating them with predefined categories. We present a
semi-supervised NER approach that starts with identifying named entities with a
small set of training data. Using the identified named entities, the word and
the context features are used to define the pattern. This pattern of each named
entity category is used as a seed pattern to identify the named entities in the
test set. Pattern scoring and tuple value score enables the generation of the
new patterns to identify the named entity categories. We have evaluated the
proposed system for English language with the dataset of tagged (IEER) and
untagged (CoNLL 2003) named entity corpus and for Tamil language with the
documents from the FIRE corpus and yield an average f-measure of 75% for both
the languages.
| no_new_dataset | 0.9463 |
1511.06838 | Stella Yu | Takuya Narihira, Damian Borth, Stella X. Yu, Karl Ni, Trevor Darrell | Mapping Images to Sentiment Adjective Noun Pairs with Factorized Neural
Nets | null | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the visual sentiment task of mapping an image to an adjective
noun pair (ANP) such as "cute baby". To capture the two-factor structure of our
ANP semantics as well as to overcome annotation noise and ambiguity, we propose
a novel factorized CNN model which learns separate representations for
adjectives and nouns but optimizes the classification performance over their
product. Our experiments on the publicly available SentiBank dataset show that
our model significantly outperforms not only independent ANP classifiers on
unseen ANPs and on retrieving images of novel ANPs, but also image captioning
models which capture word semantics from co-occurrence of natural text; the
latter turn out to be surprisingly poor at capturing the sentiment evoked by
pure visual experience. That is, our factorized ANP CNN not only trains better
from noisy labels, generalizes better to new images, but can also expands the
ANP vocabulary on its own.
| [
{
"version": "v1",
"created": "Sat, 21 Nov 2015 04:58:46 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Narihira",
"Takuya",
""
],
[
"Borth",
"Damian",
""
],
[
"Yu",
"Stella X.",
""
],
[
"Ni",
"Karl",
""
],
[
"Darrell",
"Trevor",
""
]
] | TITLE: Mapping Images to Sentiment Adjective Noun Pairs with Factorized Neural
Nets
ABSTRACT: We consider the visual sentiment task of mapping an image to an adjective
noun pair (ANP) such as "cute baby". To capture the two-factor structure of our
ANP semantics as well as to overcome annotation noise and ambiguity, we propose
a novel factorized CNN model which learns separate representations for
adjectives and nouns but optimizes the classification performance over their
product. Our experiments on the publicly available SentiBank dataset show that
our model significantly outperforms not only independent ANP classifiers on
unseen ANPs and on retrieving images of novel ANPs, but also image captioning
models which capture word semantics from co-occurrence of natural text; the
latter turn out to be surprisingly poor at capturing the sentiment evoked by
pure visual experience. That is, our factorized ANP CNN not only trains better
from noisy labels, generalizes better to new images, but can also expands the
ANP vocabulary on its own.
| no_new_dataset | 0.950869 |
1511.06853 | Yichao Xu | Yichao Xu, Hajime Nagahara, Atsushi Shimada, Rin-ichiro Taniguchi | TransCut: Transparent Object Segmentation from a Light-Field Image | 9 pages, 14 figures, 2 tables, ICCV 2015 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The segmentation of transparent objects can be very useful in computer vision
applications. However, because they borrow texture from their background and
have a similar appearance to their surroundings, transparent objects are not
handled well by regular image segmentation methods. We propose a method that
overcomes these problems using the consistency and distortion properties of a
light-field image. Graph-cut optimization is applied for the pixel labeling
problem. The light-field linearity is used to estimate the likelihood of a
pixel belonging to the transparent object or Lambertian background, and the
occlusion detector is used to find the occlusion boundary. We acquire a light
field dataset for the transparent object, and use this dataset to evaluate our
method. The results demonstrate that the proposed method successfully segments
transparent objects from the background.
| [
{
"version": "v1",
"created": "Sat, 21 Nov 2015 08:33:18 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Xu",
"Yichao",
""
],
[
"Nagahara",
"Hajime",
""
],
[
"Shimada",
"Atsushi",
""
],
[
"Taniguchi",
"Rin-ichiro",
""
]
] | TITLE: TransCut: Transparent Object Segmentation from a Light-Field Image
ABSTRACT: The segmentation of transparent objects can be very useful in computer vision
applications. However, because they borrow texture from their background and
have a similar appearance to their surroundings, transparent objects are not
handled well by regular image segmentation methods. We propose a method that
overcomes these problems using the consistency and distortion properties of a
light-field image. Graph-cut optimization is applied for the pixel labeling
problem. The light-field linearity is used to estimate the likelihood of a
pixel belonging to the transparent object or Lambertian background, and the
occlusion detector is used to find the occlusion boundary. We acquire a light
field dataset for the transparent object, and use this dataset to evaluate our
method. The results demonstrate that the proposed method successfully segments
transparent objects from the background.
| new_dataset | 0.542515 |
1511.06951 | Leslie Smith | Leslie N. Smith, Emily M. Hand, Timothy Doster | Gradual DropIn of Layers to Train Very Deep Neural Networks | null | null | null | null | cs.NE cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce the concept of dynamically growing a neural network during
training. In particular, an untrainable deep network starts as a trainable
shallow network and newly added layers are slowly, organically added during
training, thereby increasing the network's depth. This is accomplished by a new
layer, which we call DropIn. The DropIn layer starts by passing the output from
a previous layer (effectively skipping over the newly added layers), then
increasingly including units from the new layers for both feedforward and
backpropagation. We show that deep networks, which are untrainable with
conventional methods, will converge with DropIn layers interspersed in the
architecture. In addition, we demonstrate that DropIn provides regularization
during training in an analogous way as dropout. Experiments are described with
the MNIST dataset and various expanded LeNet architectures, CIFAR-10 dataset
with its architecture expanded from 3 to 11 layers, and on the ImageNet dataset
with the AlexNet architecture expanded to 13 layers and the VGG 16-layer
architecture.
| [
{
"version": "v1",
"created": "Sun, 22 Nov 2015 02:33:08 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Smith",
"Leslie N.",
""
],
[
"Hand",
"Emily M.",
""
],
[
"Doster",
"Timothy",
""
]
] | TITLE: Gradual DropIn of Layers to Train Very Deep Neural Networks
ABSTRACT: We introduce the concept of dynamically growing a neural network during
training. In particular, an untrainable deep network starts as a trainable
shallow network and newly added layers are slowly, organically added during
training, thereby increasing the network's depth. This is accomplished by a new
layer, which we call DropIn. The DropIn layer starts by passing the output from
a previous layer (effectively skipping over the newly added layers), then
increasingly including units from the new layers for both feedforward and
backpropagation. We show that deep networks, which are untrainable with
conventional methods, will converge with DropIn layers interspersed in the
architecture. In addition, we demonstrate that DropIn provides regularization
during training in an analogous way as dropout. Experiments are described with
the MNIST dataset and various expanded LeNet architectures, CIFAR-10 dataset
with its architecture expanded from 3 to 11 layers, and on the ImageNet dataset
with the AlexNet architecture expanded to 13 layers and the VGG 16-layer
architecture.
| no_new_dataset | 0.952882 |
1511.06988 | Haitian Zheng | Haitian Zheng, Yebin Liu, Mengqi Ji, Feng Wu, Lu Fang | Learning High-level Prior with Convolutional Neural Networks for
Semantic Segmentation | 9 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes a convolutional neural network that can fuse high-level
prior for semantic image segmentation. Motivated by humans' vision recognition
system, our key design is a three-layer generative structure consisting of
high-level coding, middle-level segmentation and low-level image to introduce
global prior for semantic segmentation. Based on this structure, we proposed a
generative model called conditional variational auto-encoder (CVAE) that can
build up the links behind these three layers. These important links include an
image encoder that extracts high level info from image, a segmentation encoder
that extracts high level info from segmentation, and a hybrid decoder that
outputs semantic segmentation from the high level prior and input image. We
theoretically derive the semantic segmentation as an optimization problem
parameterized by these links. Finally, the optimization problem enables us to
take advantage of state-of-the-art fully convolutional network structure for
the implementation of the above encoders and decoder. Experimental results on
several representative datasets demonstrate our supreme performance for
semantic segmentation.
| [
{
"version": "v1",
"created": "Sun, 22 Nov 2015 10:25:02 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Zheng",
"Haitian",
""
],
[
"Liu",
"Yebin",
""
],
[
"Ji",
"Mengqi",
""
],
[
"Wu",
"Feng",
""
],
[
"Fang",
"Lu",
""
]
] | TITLE: Learning High-level Prior with Convolutional Neural Networks for
Semantic Segmentation
ABSTRACT: This paper proposes a convolutional neural network that can fuse high-level
prior for semantic image segmentation. Motivated by humans' vision recognition
system, our key design is a three-layer generative structure consisting of
high-level coding, middle-level segmentation and low-level image to introduce
global prior for semantic segmentation. Based on this structure, we proposed a
generative model called conditional variational auto-encoder (CVAE) that can
build up the links behind these three layers. These important links include an
image encoder that extracts high level info from image, a segmentation encoder
that extracts high level info from segmentation, and a hybrid decoder that
outputs semantic segmentation from the high level prior and input image. We
theoretically derive the semantic segmentation as an optimization problem
parameterized by these links. Finally, the optimization problem enables us to
take advantage of state-of-the-art fully convolutional network structure for
the implementation of the above encoders and decoder. Experimental results on
several representative datasets demonstrate our supreme performance for
semantic segmentation.
| no_new_dataset | 0.948632 |
1511.07004 | Keunwoo Choi Mr | Keunwoo Choi, George Fazekas, Mark Sandler | Understanding Music Playlists | International Conference on Machine Learning (ICML) 2015, Machine
Learning for Music Discovery Workshop | null | null | null | cs.MM cs.IR | http://creativecommons.org/licenses/by/4.0/ | As music streaming services dominate the music industry, the playlist is
becoming an increasingly crucial element of music consumption. Con- sequently,
the music recommendation problem is often casted as a playlist generation prob-
lem. Better understanding of the playlist is there- fore necessary for
developing better playlist gen- eration algorithms. In this work, we analyse
two playlist datasets to investigate some com- monly assumed hypotheses about
playlists. Our findings indicate that deeper understanding of playlists is
needed to provide better prior infor- mation and improve machine learning
algorithms in the design of recommendation systems.
| [
{
"version": "v1",
"created": "Sun, 22 Nov 2015 12:33:08 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Choi",
"Keunwoo",
""
],
[
"Fazekas",
"George",
""
],
[
"Sandler",
"Mark",
""
]
] | TITLE: Understanding Music Playlists
ABSTRACT: As music streaming services dominate the music industry, the playlist is
becoming an increasingly crucial element of music consumption. Con- sequently,
the music recommendation problem is often casted as a playlist generation prob-
lem. Better understanding of the playlist is there- fore necessary for
developing better playlist gen- eration algorithms. In this work, we analyse
two playlist datasets to investigate some com- monly assumed hypotheses about
playlists. Our findings indicate that deeper understanding of playlists is
needed to provide better prior infor- mation and improve machine learning
algorithms in the design of recommendation systems.
| no_new_dataset | 0.95222 |
1511.07017 | Sudhakar Singh | Sudhakar Singh, Rakhi Garg, P. K. Mishra | Performance Analysis of Apriori Algorithm with Different Data Structures
on Hadoop Cluster | 2009-2015 International Journal of Computer Applications,
FCS(Foundation of Computer Science) | null | 10.5120/ijca2015906632 | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mining frequent itemsets from massive datasets is always being a most
important problem of data mining. Apriori is the most popular and simplest
algorithm for frequent itemset mining. To enhance the efficiency and
scalability of Apriori, a number of algorithms have been proposed addressing
the design of efficient data structures, minimizing database scan and parallel
and distributed processing. MapReduce is the emerging parallel and distributed
technology to process big datasets on Hadoop Cluster. To mine big datasets it
is essential to re-design the data mining algorithm on this new paradigm. In
this paper, we implement three variations of Apriori algorithm using data
structures hash tree, trie and hash table trie i.e. trie with hash technique on
MapReduce paradigm. We emphasize and investigate the significance of these
three data structures for Apriori algorithm on Hadoop cluster, which has not
been given attention yet. Experiments are carried out on both real life and
synthetic datasets which shows that hash table trie data structures performs
far better than trie and hash tree in terms of execution time. Moreover the
performance in case of hash tree becomes worst.
| [
{
"version": "v1",
"created": "Sun, 22 Nov 2015 14:40:06 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Singh",
"Sudhakar",
""
],
[
"Garg",
"Rakhi",
""
],
[
"Mishra",
"P. K.",
""
]
] | TITLE: Performance Analysis of Apriori Algorithm with Different Data Structures
on Hadoop Cluster
ABSTRACT: Mining frequent itemsets from massive datasets is always being a most
important problem of data mining. Apriori is the most popular and simplest
algorithm for frequent itemset mining. To enhance the efficiency and
scalability of Apriori, a number of algorithms have been proposed addressing
the design of efficient data structures, minimizing database scan and parallel
and distributed processing. MapReduce is the emerging parallel and distributed
technology to process big datasets on Hadoop Cluster. To mine big datasets it
is essential to re-design the data mining algorithm on this new paradigm. In
this paper, we implement three variations of Apriori algorithm using data
structures hash tree, trie and hash table trie i.e. trie with hash technique on
MapReduce paradigm. We emphasize and investigate the significance of these
three data structures for Apriori algorithm on Hadoop cluster, which has not
been given attention yet. Experiments are carried out on both real life and
synthetic datasets which shows that hash table trie data structures performs
far better than trie and hash tree in terms of execution time. Moreover the
performance in case of hash tree becomes worst.
| no_new_dataset | 0.949342 |
1511.07063 | Ning Zhang | Ning Zhang, Evan Shelhamer, Yang Gao, Trevor Darrell | Fine-grained pose prediction, normalization, and recognition | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pose variation and subtle differences in appearance are key challenges to
fine-grained classification. While deep networks have markedly improved general
recognition, many approaches to fine-grained recognition rely on anchoring
networks to parts for better accuracy. Identifying parts to find correspondence
discounts pose variation so that features can be tuned to appearance. To this
end previous methods have examined how to find parts and extract
pose-normalized features. These methods have generally separated fine-grained
recognition into stages which first localize parts using hand-engineered and
coarsely-localized proposal features, and then separately learn deep
descriptors centered on inferred part positions. We unify these steps in an
end-to-end trainable network supervised by keypoint locations and class labels
that localizes parts by a fully convolutional network to focus the learning of
feature representations for the fine-grained classification task. Experiments
on the popular CUB200 dataset show that our method is state-of-the-art and
suggest a continuing role for strong supervision.
| [
{
"version": "v1",
"created": "Sun, 22 Nov 2015 20:32:45 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Zhang",
"Ning",
""
],
[
"Shelhamer",
"Evan",
""
],
[
"Gao",
"Yang",
""
],
[
"Darrell",
"Trevor",
""
]
] | TITLE: Fine-grained pose prediction, normalization, and recognition
ABSTRACT: Pose variation and subtle differences in appearance are key challenges to
fine-grained classification. While deep networks have markedly improved general
recognition, many approaches to fine-grained recognition rely on anchoring
networks to parts for better accuracy. Identifying parts to find correspondence
discounts pose variation so that features can be tuned to appearance. To this
end previous methods have examined how to find parts and extract
pose-normalized features. These methods have generally separated fine-grained
recognition into stages which first localize parts using hand-engineered and
coarsely-localized proposal features, and then separately learn deep
descriptors centered on inferred part positions. We unify these steps in an
end-to-end trainable network supervised by keypoint locations and class labels
that localizes parts by a fully convolutional network to focus the learning of
feature representations for the fine-grained classification task. Experiments
on the popular CUB200 dataset show that our method is state-of-the-art and
suggest a continuing role for strong supervision.
| no_new_dataset | 0.947769 |
1511.07340 | Henry WJ Reeve | Henry W J Reeve and Gavin Brown | Modular Autoencoders for Ensemble Feature Extraction | 18 pages, 8 figures, to appear in a special issue of The Journal Of
Machine Learning Research (vol.44, Dec 2015) | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce the concept of a Modular Autoencoder (MAE), capable of learning
a set of diverse but complementary representations from unlabelled data, that
can later be used for supervised tasks. The learning of the representations is
controlled by a trade off parameter, and we show on six benchmark datasets the
optimum lies between two extremes: a set of smaller, independent autoencoders
each with low capacity, versus a single monolithic encoding, outperforming an
appropriate baseline. In the present paper we explore the special case of
linear MAE, and derive an SVD-based algorithm which converges several orders of
magnitude faster than gradient descent.
| [
{
"version": "v1",
"created": "Mon, 23 Nov 2015 17:51:18 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Reeve",
"Henry W J",
""
],
[
"Brown",
"Gavin",
""
]
] | TITLE: Modular Autoencoders for Ensemble Feature Extraction
ABSTRACT: We introduce the concept of a Modular Autoencoder (MAE), capable of learning
a set of diverse but complementary representations from unlabelled data, that
can later be used for supervised tasks. The learning of the representations is
controlled by a trade off parameter, and we show on six benchmark datasets the
optimum lies between two extremes: a set of smaller, independent autoencoders
each with low capacity, versus a single monolithic encoding, outperforming an
appropriate baseline. In the present paper we explore the special case of
linear MAE, and derive an SVD-based algorithm which converges several orders of
magnitude faster than gradient descent.
| no_new_dataset | 0.946597 |
1511.07353 | Ahmed Bin Shafaat Ahmed Bin Shafaat | Kamran Shaukat, Nayyer Masood, Ahmed Bin Shafaat, Kamran Jabbar,
Hassan Shabbir and Shakir Shabbir | Dengue Fever in Perspective of Clustering Algorithms | null | null | null | null | cs.CY | http://creativecommons.org/licenses/by/4.0/ | Dengue fever is a disease which is transmitted and caused by Aedes Aegypti
mosquitos. Dengue has become a serious health issue in all over the world
especially in those countries who are situated in tropical or subtropical
regions because rain is an important factor for growth and increase in the
population of dengue transmitting mosquitos. For a long time, data mining
algorithms have been used by the scientists for the diagnosis and prognosis of
different diseases which includes dengue as well. This was a study to analyses
the attack of dengue fever in different areas of district Jhelum, Pakistan in
2011. As per our knowledge, we are unaware of any kind of research study in the
area of district Jhelum for diagnosis or analysis of dengue fever. According to
our information, we are the first one researching and analyzing dengue fever in
this specific area. Dataset was obtained from the office of Executive District
Officer EDO (health) District Jhelum. We applied DBSCAN algorithm for the
clustering of dengue fever. First we showed overall behavior of dengue in the
district Jhelum. Then we explained dengue fever at tehsil level with the help
of geographical pictures. After that we have elaborated comparison of different
clustering algorithms with the help of graphs based on our dataset. Those
algorithms include k-means, K-mediods, DBSCAN and OPTICS.
| [
{
"version": "v1",
"created": "Mon, 23 Nov 2015 18:27:21 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Shaukat",
"Kamran",
""
],
[
"Masood",
"Nayyer",
""
],
[
"Shafaat",
"Ahmed Bin",
""
],
[
"Jabbar",
"Kamran",
""
],
[
"Shabbir",
"Hassan",
""
],
[
"Shabbir",
"Shakir",
""
]
] | TITLE: Dengue Fever in Perspective of Clustering Algorithms
ABSTRACT: Dengue fever is a disease which is transmitted and caused by Aedes Aegypti
mosquitos. Dengue has become a serious health issue in all over the world
especially in those countries who are situated in tropical or subtropical
regions because rain is an important factor for growth and increase in the
population of dengue transmitting mosquitos. For a long time, data mining
algorithms have been used by the scientists for the diagnosis and prognosis of
different diseases which includes dengue as well. This was a study to analyses
the attack of dengue fever in different areas of district Jhelum, Pakistan in
2011. As per our knowledge, we are unaware of any kind of research study in the
area of district Jhelum for diagnosis or analysis of dengue fever. According to
our information, we are the first one researching and analyzing dengue fever in
this specific area. Dataset was obtained from the office of Executive District
Officer EDO (health) District Jhelum. We applied DBSCAN algorithm for the
clustering of dengue fever. First we showed overall behavior of dengue in the
district Jhelum. Then we explained dengue fever at tehsil level with the help
of geographical pictures. After that we have elaborated comparison of different
clustering algorithms with the help of graphs based on our dataset. Those
algorithms include k-means, K-mediods, DBSCAN and OPTICS.
| no_new_dataset | 0.927953 |
1511.07357 | Sariel Har-Peled | Sariel Har-Peled and Sepideh Mahabadi | Proximity in the Age of Distraction: Robust Approximate Nearest Neighbor
Search | null | null | null | null | cs.CG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a new variant of the nearest neighbor search problem, which
allows for some coordinates of the dataset to be arbitrarily corrupted or
unknown. Formally, given a dataset of $n$ points $P=\{ x_1,\ldots, x_n\}$ in
high-dimensions, and a parameter $k$, the goal is to preprocess the dataset,
such that given a query point $q$, one can compute quickly a point $x \in P$,
such that the distance of the query to the point $x$ is minimized, when
ignoring the "optimal" $k$ coordinates. Note, that the coordinates being
ignored are a function of both the query point and the point returned.
We present a general reduction from this problem to answering ANN queries,
which is similar in spirit to LSH (locality sensitive hashing) [IM98].
Specifically, we give a sampling technique which achieves a bi-criterion
approximation for this problem. If the distance to the nearest neighbor after
ignoring $k$ coordinates is $r$, the data-structure returns a point that is
within a distance of $O(r)$ after ignoring $O(k)$ coordinates. We also present
other applications and further extensions and refinements of the above result.
The new data-structures are simple and (arguably) elegant, and should be
practical -- specifically, all bounds are polynomial in all relevant parameters
(including the dimension of the space, and the robustness parameter $k$).
| [
{
"version": "v1",
"created": "Mon, 23 Nov 2015 18:45:12 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Har-Peled",
"Sariel",
""
],
[
"Mahabadi",
"Sepideh",
""
]
] | TITLE: Proximity in the Age of Distraction: Robust Approximate Nearest Neighbor
Search
ABSTRACT: We introduce a new variant of the nearest neighbor search problem, which
allows for some coordinates of the dataset to be arbitrarily corrupted or
unknown. Formally, given a dataset of $n$ points $P=\{ x_1,\ldots, x_n\}$ in
high-dimensions, and a parameter $k$, the goal is to preprocess the dataset,
such that given a query point $q$, one can compute quickly a point $x \in P$,
such that the distance of the query to the point $x$ is minimized, when
ignoring the "optimal" $k$ coordinates. Note, that the coordinates being
ignored are a function of both the query point and the point returned.
We present a general reduction from this problem to answering ANN queries,
which is similar in spirit to LSH (locality sensitive hashing) [IM98].
Specifically, we give a sampling technique which achieves a bi-criterion
approximation for this problem. If the distance to the nearest neighbor after
ignoring $k$ coordinates is $r$, the data-structure returns a point that is
within a distance of $O(r)$ after ignoring $O(k)$ coordinates. We also present
other applications and further extensions and refinements of the above result.
The new data-structures are simple and (arguably) elegant, and should be
practical -- specifically, all bounds are polynomial in all relevant parameters
(including the dimension of the space, and the robustness parameter $k$).
| no_new_dataset | 0.938067 |
1511.07397 | Gabriele Farina | Gabriele Farina, Nicola Gatti | Ad auctions and cascade model: GSP inefficiency and algorithms | AAAI16, to appear | null | null | null | cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The design of the best economic mechanism for Sponsored Search Auctions
(SSAs) is a central task in computational mechanism design/game theory. Two
open questions concern the adoption of user models more accurate than that one
currently used and the choice between Generalized Second Price auction (GSP)
and Vickrey-Clark-Groves mechanism (VCG). In this paper, we provide some
contributions to answer these questions. We study Price of Anarchy (PoA) and
Price of Stability (PoS) over social welfare and auctioneer's revenue of GSP
w.r.t. the VCG when the users follow the famous cascade model. Furthermore, we
provide exact, randomized, and approximate algorithms, showing that in
real-world settings (Yahoo! Webscope A3 dataset, 10 available slots) optimal
allocations can be found in less than 1s with up to 1000 ads, and can be
approximated in less than 20ms even with more than 1000 ads with an average
accuracy greater than 99%.
| [
{
"version": "v1",
"created": "Mon, 23 Nov 2015 20:19:55 GMT"
}
] | 2015-11-24T00:00:00 | [
[
"Farina",
"Gabriele",
""
],
[
"Gatti",
"Nicola",
""
]
] | TITLE: Ad auctions and cascade model: GSP inefficiency and algorithms
ABSTRACT: The design of the best economic mechanism for Sponsored Search Auctions
(SSAs) is a central task in computational mechanism design/game theory. Two
open questions concern the adoption of user models more accurate than that one
currently used and the choice between Generalized Second Price auction (GSP)
and Vickrey-Clark-Groves mechanism (VCG). In this paper, we provide some
contributions to answer these questions. We study Price of Anarchy (PoA) and
Price of Stability (PoS) over social welfare and auctioneer's revenue of GSP
w.r.t. the VCG when the users follow the famous cascade model. Furthermore, we
provide exact, randomized, and approximate algorithms, showing that in
real-world settings (Yahoo! Webscope A3 dataset, 10 available slots) optimal
allocations can be found in less than 1s with up to 1000 ads, and can be
approximated in less than 20ms even with more than 1000 ads with an average
accuracy greater than 99%.
| no_new_dataset | 0.945399 |
1412.5474 | Jonghoon Jin | Jonghoon Jin, Aysegul Dundar, Eugenio Culurciello | Flattened Convolutional Neural Networks for Feedforward Acceleration | International Conference on Learning Representations (ICLR) 2015 | null | null | null | cs.NE cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present flattened convolutional neural networks that are designed for fast
feedforward execution. The redundancy of the parameters, especially weights of
the convolutional filters in convolutional neural networks has been extensively
studied and different heuristics have been proposed to construct a low rank
basis of the filters after training. In this work, we train flattened networks
that consist of consecutive sequence of one-dimensional filters across all
directions in 3D space to obtain comparable performance as conventional
convolutional networks. We tested flattened model on different datasets and
found that the flattened layer can effectively substitute for the 3D filters
without loss of accuracy. The flattened convolution pipelines provide around
two times speed-up during feedforward pass compared to the baseline model due
to the significant reduction of learning parameters. Furthermore, the proposed
method does not require efforts in manual tuning or post processing once the
model is trained.
| [
{
"version": "v1",
"created": "Wed, 17 Dec 2014 16:48:54 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Feb 2015 20:36:05 GMT"
},
{
"version": "v3",
"created": "Mon, 6 Apr 2015 01:40:08 GMT"
},
{
"version": "v4",
"created": "Fri, 20 Nov 2015 05:50:23 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Jin",
"Jonghoon",
""
],
[
"Dundar",
"Aysegul",
""
],
[
"Culurciello",
"Eugenio",
""
]
] | TITLE: Flattened Convolutional Neural Networks for Feedforward Acceleration
ABSTRACT: We present flattened convolutional neural networks that are designed for fast
feedforward execution. The redundancy of the parameters, especially weights of
the convolutional filters in convolutional neural networks has been extensively
studied and different heuristics have been proposed to construct a low rank
basis of the filters after training. In this work, we train flattened networks
that consist of consecutive sequence of one-dimensional filters across all
directions in 3D space to obtain comparable performance as conventional
convolutional networks. We tested flattened model on different datasets and
found that the flattened layer can effectively substitute for the 3D filters
without loss of accuracy. The flattened convolution pipelines provide around
two times speed-up during feedforward pass compared to the baseline model due
to the significant reduction of learning parameters. Furthermore, the proposed
method does not require efforts in manual tuning or post processing once the
model is trained.
| no_new_dataset | 0.950869 |
1412.7957 | Sezer Karaoglu | Sezer Karaoglu, Yang Liu, Theo Gevers | Detect2Rank : Combining Object Detectors Using Learning to Rank | null | null | 10.1109/TIP.2015.2499702 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object detection is an important research area in the field of computer
vision. Many detection algorithms have been proposed. However, each object
detector relies on specific assumptions of the object appearance and imaging
conditions. As a consequence, no algorithm can be considered as universal. With
the large variety of object detectors, the subsequent question is how to select
and combine them.
In this paper, we propose a framework to learn how to combine object
detectors. The proposed method uses (single) detectors like DPM, CN and EES,
and exploits their correlation by high level contextual features to yield a
combined detection list.
Experiments on the PASCAL VOC07 and VOC10 datasets show that the proposed
method significantly outperforms single object detectors, DPM (8.4%), CN (6.8%)
and EES (17.0%) on VOC07 and DPM (6.5%), CN (5.5%) and EES (16.2%) on VOC10.
| [
{
"version": "v1",
"created": "Fri, 26 Dec 2014 16:46:52 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Karaoglu",
"Sezer",
""
],
[
"Liu",
"Yang",
""
],
[
"Gevers",
"Theo",
""
]
] | TITLE: Detect2Rank : Combining Object Detectors Using Learning to Rank
ABSTRACT: Object detection is an important research area in the field of computer
vision. Many detection algorithms have been proposed. However, each object
detector relies on specific assumptions of the object appearance and imaging
conditions. As a consequence, no algorithm can be considered as universal. With
the large variety of object detectors, the subsequent question is how to select
and combine them.
In this paper, we propose a framework to learn how to combine object
detectors. The proposed method uses (single) detectors like DPM, CN and EES,
and exploits their correlation by high level contextual features to yield a
combined detection list.
Experiments on the PASCAL VOC07 and VOC10 datasets show that the proposed
method significantly outperforms single object detectors, DPM (8.4%), CN (6.8%)
and EES (17.0%) on VOC07 and DPM (6.5%), CN (5.5%) and EES (16.2%) on VOC10.
| no_new_dataset | 0.951818 |
1501.05928 | John L. Rubinstein | Jianhua Zhao, Marcus A. Brubaker, Samir Benlekbir, John L. Rubinstein | Description and comparison of algorithms for correcting anisotropic
magnification in cryo-EM images | 9 pages, 4 figures | J Struct Biol. 2015 Nov;192(2):209-15. doi:
10.1016/j.jsb.2015.06.014 | null | null | physics.ins-det q-bio.BM q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Single particle electron cryomicroscopy (cryo-EM) allows for structures of
proteins and protein complexes to be determined from images of non-crystalline
specimens. Cryo-EM data analysis requires electron microscope images of
randomly oriented ice-embedded protein particles to be rotated and translated
to allow for coherent averaging when calculating three-dimensional (3D)
structures. Rotation of 2D images is usually done with the assumption that the
magnification of the electron microscope is the same in all directions.
However, due to electron optical aberrations, this condition is not met with
some electron microscopes when used with the settings necessary for cryo-EM
with a direct detector device (DDD) camera. Correction of images by linear
interpolation in real space has allowed high-resolution structures to be
calculated from cryo-EM images for symmetric particles. Here we describe and
compare a simple real space method, a simple Fourier space method, and a
somewhat more sophisticated Fourier space method to correct images for a
measured anisotropy in magnification. Further, anisotropic magnification causes
contrast transfer function (CTF) parameters estimated from image power spectra
to have an apparent systematic astigmatism. To address this problem we develop
an approach to adjust CTF parameters measured from distorted images so that
they can be used with corrected images. The effect of anisotropic magnification
on CTF parameters provides a simple way of detecting magnification anisotropy
in cryo-EM datasets.
| [
{
"version": "v1",
"created": "Fri, 23 Jan 2015 19:51:40 GMT"
},
{
"version": "v2",
"created": "Tue, 27 Jan 2015 22:56:39 GMT"
},
{
"version": "v3",
"created": "Fri, 7 Aug 2015 21:06:45 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Zhao",
"Jianhua",
""
],
[
"Brubaker",
"Marcus A.",
""
],
[
"Benlekbir",
"Samir",
""
],
[
"Rubinstein",
"John L.",
""
]
] | TITLE: Description and comparison of algorithms for correcting anisotropic
magnification in cryo-EM images
ABSTRACT: Single particle electron cryomicroscopy (cryo-EM) allows for structures of
proteins and protein complexes to be determined from images of non-crystalline
specimens. Cryo-EM data analysis requires electron microscope images of
randomly oriented ice-embedded protein particles to be rotated and translated
to allow for coherent averaging when calculating three-dimensional (3D)
structures. Rotation of 2D images is usually done with the assumption that the
magnification of the electron microscope is the same in all directions.
However, due to electron optical aberrations, this condition is not met with
some electron microscopes when used with the settings necessary for cryo-EM
with a direct detector device (DDD) camera. Correction of images by linear
interpolation in real space has allowed high-resolution structures to be
calculated from cryo-EM images for symmetric particles. Here we describe and
compare a simple real space method, a simple Fourier space method, and a
somewhat more sophisticated Fourier space method to correct images for a
measured anisotropy in magnification. Further, anisotropic magnification causes
contrast transfer function (CTF) parameters estimated from image power spectra
to have an apparent systematic astigmatism. To address this problem we develop
an approach to adjust CTF parameters measured from distorted images so that
they can be used with corrected images. The effect of anisotropic magnification
on CTF parameters provides a simple way of detecting magnification anisotropy
in cryo-EM datasets.
| no_new_dataset | 0.954308 |
1511.05688 | Ameen Eetemadi | Ameen Eetemadi, Ilias Tagkopoulos | A Distribution Adaptive Framework for Prediction Interval Estimation
Using Nominal Variables | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Proposed methods for prediction interval estimation so far focus on cases
where input variables are numerical. In datasets with solely nominal input
variables, we observe records with the exact same input $x^u$, but different
real valued outputs due to the inherent noise in the system. Existing
prediction interval estimation methods do not use representations that can
accurately model such inherent noise in the case of nominal inputs. We propose
a new prediction interval estimation method tailored for this type of data,
which is prevalent in biology and medicine. We call this method Distribution
Adaptive Prediction Interval Estimation given Nominal inputs (DAPIEN) and has
four main phases. First, we select a distribution function that can best
represent the inherent noise of the system for all unique inputs. Then we infer
the parameters $\theta_i$ (e.g. $\theta_i=[mean_i, variance_i]$) of the
selected distribution function for all unique input vectors $x^u_i$ and
generate a new corresponding training set using pairs of $x^u_i, \theta_i$.
III). Then, we train a model to predict $\theta$ given a new $x_u$. Finally, we
calculate the prediction interval for a new sample using the inverse of the
cumulative distribution function once the parameters $\theta$ is predicted by
the trained model. We compared DAPIEN to the commonly used Bootstrap method on
three synthetic datasets. Our results show that DAPIEN provides tighter
prediction intervals while preserving the requested coverage when compared to
Bootstrap. This work can facilitate broader usage of regression methods in
medicine and biology where it is necessary to provide tight prediction
intervals while preserving coverage when input variables are nominal.
| [
{
"version": "v1",
"created": "Wed, 18 Nov 2015 08:13:35 GMT"
},
{
"version": "v2",
"created": "Fri, 20 Nov 2015 08:12:23 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Eetemadi",
"Ameen",
""
],
[
"Tagkopoulos",
"Ilias",
""
]
] | TITLE: A Distribution Adaptive Framework for Prediction Interval Estimation
Using Nominal Variables
ABSTRACT: Proposed methods for prediction interval estimation so far focus on cases
where input variables are numerical. In datasets with solely nominal input
variables, we observe records with the exact same input $x^u$, but different
real valued outputs due to the inherent noise in the system. Existing
prediction interval estimation methods do not use representations that can
accurately model such inherent noise in the case of nominal inputs. We propose
a new prediction interval estimation method tailored for this type of data,
which is prevalent in biology and medicine. We call this method Distribution
Adaptive Prediction Interval Estimation given Nominal inputs (DAPIEN) and has
four main phases. First, we select a distribution function that can best
represent the inherent noise of the system for all unique inputs. Then we infer
the parameters $\theta_i$ (e.g. $\theta_i=[mean_i, variance_i]$) of the
selected distribution function for all unique input vectors $x^u_i$ and
generate a new corresponding training set using pairs of $x^u_i, \theta_i$.
III). Then, we train a model to predict $\theta$ given a new $x_u$. Finally, we
calculate the prediction interval for a new sample using the inverse of the
cumulative distribution function once the parameters $\theta$ is predicted by
the trained model. We compared DAPIEN to the commonly used Bootstrap method on
three synthetic datasets. Our results show that DAPIEN provides tighter
prediction intervals while preserving the requested coverage when compared to
Bootstrap. This work can facilitate broader usage of regression methods in
medicine and biology where it is necessary to provide tight prediction
intervals while preserving coverage when input variables are nominal.
| no_new_dataset | 0.948585 |
1511.06379 | Richard Searle Dr | Richard Searle, Megan Bingham-Walker | Dynamic Adaptive Network Intelligence | 8 pages, 2 figures, 3 tables, ICLR 2016 conference paper submission | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate representational learning of both the explicit and implicit
relationships within data is critical to the ability of machines to perform
more complex and abstract reasoning tasks. We describe the efficient weakly
supervised learning of such inferences by our Dynamic Adaptive Network
Intelligence (DANI) model. We report state-of-the-art results for DANI over
question answering tasks in the bAbI dataset that have proved difficult for
contemporary approaches to learning representation (Weston et al., 2015).
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 21:07:27 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Searle",
"Richard",
""
],
[
"Bingham-Walker",
"Megan",
""
]
] | TITLE: Dynamic Adaptive Network Intelligence
ABSTRACT: Accurate representational learning of both the explicit and implicit
relationships within data is critical to the ability of machines to perform
more complex and abstract reasoning tasks. We describe the efficient weakly
supervised learning of such inferences by our Dynamic Adaptive Network
Intelligence (DANI) model. We report state-of-the-art results for DANI over
question answering tasks in the bAbI dataset that have proved difficult for
contemporary approaches to learning representation (Weston et al., 2015).
| no_new_dataset | 0.942771 |
1511.06438 | Danushka Bollegala | Danushka Bollegala, Alsuhaibani Mohammed, Takanori Maehara, Ken-ichi
Kawarabayashi | Joint Word Representation Learning using a Corpus and a Semantic Lexicon | Accepted to AAAI-2016 | Proceedings of the AAAI 2016 | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Methods for learning word representations using large text corpora have
received much attention lately due to their impressive performance in numerous
natural language processing (NLP) tasks such as, semantic similarity
measurement, and word analogy detection. Despite their success, these
data-driven word representation learning methods do not consider the rich
semantic relational structure between words in a co-occurring context. On the
other hand, already much manual effort has gone into the construction of
semantic lexicons such as the WordNet that represent the meanings of words by
defining the various relationships that exist among the words in a language. We
consider the question, can we improve the word representations learnt using a
corpora by integrating the knowledge from semantic lexicons?. For this purpose,
we propose a joint word representation learning method that simultaneously
predicts the co-occurrences of two words in a sentence subject to the
relational constrains given by the semantic lexicon. We use relations that
exist between words in the lexicon to regularize the word representations
learnt from the corpus. Our proposed method statistically significantly
outperforms previously proposed methods for incorporating semantic lexicons
into word representations on several benchmark datasets for semantic similarity
and word analogy.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 22:58:10 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Bollegala",
"Danushka",
""
],
[
"Mohammed",
"Alsuhaibani",
""
],
[
"Maehara",
"Takanori",
""
],
[
"Kawarabayashi",
"Ken-ichi",
""
]
] | TITLE: Joint Word Representation Learning using a Corpus and a Semantic Lexicon
ABSTRACT: Methods for learning word representations using large text corpora have
received much attention lately due to their impressive performance in numerous
natural language processing (NLP) tasks such as, semantic similarity
measurement, and word analogy detection. Despite their success, these
data-driven word representation learning methods do not consider the rich
semantic relational structure between words in a co-occurring context. On the
other hand, already much manual effort has gone into the construction of
semantic lexicons such as the WordNet that represent the meanings of words by
defining the various relationships that exist among the words in a language. We
consider the question, can we improve the word representations learnt using a
corpora by integrating the knowledge from semantic lexicons?. For this purpose,
we propose a joint word representation learning method that simultaneously
predicts the co-occurrences of two words in a sentence subject to the
relational constrains given by the semantic lexicon. We use relations that
exist between words in the lexicon to regularize the word representations
learnt from the corpus. Our proposed method statistically significantly
outperforms previously proposed methods for incorporating semantic lexicons
into word representations on several benchmark datasets for semantic similarity
and word analogy.
| no_new_dataset | 0.944689 |
1511.06452 | Hyun Oh Song | Hyun Oh Song, Yu Xiang, Stefanie Jegelka, Silvio Savarese | Deep Metric Learning via Lifted Structured Feature Embedding | 11 pages | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning the distance metric between pairs of examples is of great importance
for learning and visual recognition. With the remarkable success from the state
of the art convolutional neural networks, recent works have shown promising
results on discriminatively training the networks to learn semantic feature
embeddings where similar examples are mapped close to each other and dissimilar
examples are mapped farther apart. In this paper, we describe an algorithm for
taking full advantage of the training batches in the neural network training by
lifting the vector of pairwise distances within the batch to the matrix of
pairwise distances. This step enables the algorithm to learn the state of the
art feature embedding by optimizing a novel structured prediction objective on
the lifted problem. Additionally, we collected Online Products dataset: 120k
images of 23k classes of online products for metric learning. Our experiments
on the CUB-200-2011, CARS196, and Online Products datasets demonstrate
significant improvement over existing deep feature embedding methods on all
experimented embedding sizes with the GoogLeNet network.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 23:41:11 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Song",
"Hyun Oh",
""
],
[
"Xiang",
"Yu",
""
],
[
"Jegelka",
"Stefanie",
""
],
[
"Savarese",
"Silvio",
""
]
] | TITLE: Deep Metric Learning via Lifted Structured Feature Embedding
ABSTRACT: Learning the distance metric between pairs of examples is of great importance
for learning and visual recognition. With the remarkable success from the state
of the art convolutional neural networks, recent works have shown promising
results on discriminatively training the networks to learn semantic feature
embeddings where similar examples are mapped close to each other and dissimilar
examples are mapped farther apart. In this paper, we describe an algorithm for
taking full advantage of the training batches in the neural network training by
lifting the vector of pairwise distances within the batch to the matrix of
pairwise distances. This step enables the algorithm to learn the state of the
art feature embedding by optimizing a novel structured prediction objective on
the lifted problem. Additionally, we collected Online Products dataset: 120k
images of 23k classes of online products for metric learning. Our experiments
on the CUB-200-2011, CARS196, and Online Products datasets demonstrate
significant improvement over existing deep feature embedding methods on all
experimented embedding sizes with the GoogLeNet network.
| new_dataset | 0.963265 |
1511.06463 | Sucheta Soundarajan | Sucheta Soundarajan, Tina Eliassi-Rad, Brian Gallagher, Ali Pinar | MaxOutProbe: An Algorithm for Increasing the Size of Partially Observed
Networks | NIPS Workshop on Networks in the Social and Information Sciences | null | null | LLNL-CONF-677677 | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Networked representations of real-world phenomena are often partially
observed, which lead to incomplete networks. Analysis of such incomplete
networks can lead to skewed results. We examine the following problem: given an
incomplete network, which $b$ nodes should be probed to bring the largest
number of new nodes into the observed network? Many graph-mining tasks require
having observed a considerable amount of the network. Examples include
community discovery, belief propagation, influence maximization, etc. For
instance, consider someone who has observed a portion (say 1%) of the Twitter
retweet network via random tweet sampling. She wants to estimate the size of
the largest connected component of the fully observed retweet network. To
improve her estimate, how should she use her limited budget to reduce the
incompleteness of the network? In this work, we propose a novel algorithm,
called MaxOutProbe, which uses a budget $b$ (on nodes probed) to increase the
size of the observed network in terms of the number of nodes. Our experiments,
across a range of datasets and conditions, demonstrate the advantages of
MaxOutProbe over existing methods.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 00:26:27 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Soundarajan",
"Sucheta",
""
],
[
"Eliassi-Rad",
"Tina",
""
],
[
"Gallagher",
"Brian",
""
],
[
"Pinar",
"Ali",
""
]
] | TITLE: MaxOutProbe: An Algorithm for Increasing the Size of Partially Observed
Networks
ABSTRACT: Networked representations of real-world phenomena are often partially
observed, which lead to incomplete networks. Analysis of such incomplete
networks can lead to skewed results. We examine the following problem: given an
incomplete network, which $b$ nodes should be probed to bring the largest
number of new nodes into the observed network? Many graph-mining tasks require
having observed a considerable amount of the network. Examples include
community discovery, belief propagation, influence maximization, etc. For
instance, consider someone who has observed a portion (say 1%) of the Twitter
retweet network via random tweet sampling. She wants to estimate the size of
the largest connected component of the fully observed retweet network. To
improve her estimate, how should she use her limited budget to reduce the
incompleteness of the network? In this work, we propose a novel algorithm,
called MaxOutProbe, which uses a budget $b$ (on nodes probed) to increase the
size of the observed network in terms of the number of nodes. Our experiments,
across a range of datasets and conditions, demonstrate the advantages of
MaxOutProbe over existing methods.
| no_new_dataset | 0.947235 |
1511.06523 | Shuo Yang | Shuo Yang, Ping Luo, Chen Change Loy, and Xiaoou Tang | WIDER FACE: A Face Detection Benchmark | 12 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Face detection is one of the most studied topics in the computer vision
community. Much of the progresses have been made by the availability of face
detection benchmark datasets. We show that there is a gap between current face
detection performance and the real world requirements. To facilitate future
face detection research, we introduce the WIDER FACE dataset, which is 10 times
larger than existing datasets. The dataset contains rich annotations, including
occlusions, poses, event categories, and face bounding boxes. Faces in the
proposed dataset are extremely challenging due to large variations in scale,
pose and occlusion, as shown in Fig. 1. Furthermore, we show that WIDER FACE
dataset is an effective training source for face detection. We benchmark
several representative detection systems, providing an overview of
state-of-the-art performance and propose a solution to deal with large scale
variation. Finally, we discuss common failure cases that worth to be further
investigated. Dataset can be downloaded at:
mmlab.ie.cuhk.edu.hk/projects/WIDERFace
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 08:33:57 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Yang",
"Shuo",
""
],
[
"Luo",
"Ping",
""
],
[
"Loy",
"Chen Change",
""
],
[
"Tang",
"Xiaoou",
""
]
] | TITLE: WIDER FACE: A Face Detection Benchmark
ABSTRACT: Face detection is one of the most studied topics in the computer vision
community. Much of the progresses have been made by the availability of face
detection benchmark datasets. We show that there is a gap between current face
detection performance and the real world requirements. To facilitate future
face detection research, we introduce the WIDER FACE dataset, which is 10 times
larger than existing datasets. The dataset contains rich annotations, including
occlusions, poses, event categories, and face bounding boxes. Faces in the
proposed dataset are extremely challenging due to large variations in scale,
pose and occlusion, as shown in Fig. 1. Furthermore, we show that WIDER FACE
dataset is an effective training source for face detection. We benchmark
several representative detection systems, providing an overview of
state-of-the-art performance and propose a solution to deal with large scale
variation. Finally, we discuss common failure cases that worth to be further
investigated. Dataset can be downloaded at:
mmlab.ie.cuhk.edu.hk/projects/WIDERFace
| new_dataset | 0.959988 |
1511.06554 | Li Li | Li Li, Tegawend\'e F. Bissyand\'e, Jacques Klein, Yves Le Traon | An Investigation into the Use of Common Libraries in Android Apps | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The packaging model of Android apps requires the entire code necessary for
the execution of an app to be shipped into one single apk file. Thus, an
analysis of Android apps often visits code which is not part of the
functionality delivered by the app. Such code is often contributed by the
common libraries which are used pervasively by all apps. Unfortunately, Android
analyses, e.g., for piggybacking detection and malware detection, can produce
inaccurate results if they do not take into account the case of library code,
which constitute noise in app features. Despite some efforts on investigating
Android libraries, the momentum of Android research has not yet produced a
complete set of common libraries to further support in-depth analysis of
Android apps. In this paper, we leverage a dataset of about 1.5 million apps
from Google Play to harvest potential common libraries, including advertisement
libraries. With several steps of refinements, we finally collect by far the
largest set of 1,113 libraries supporting common functionalities and 240
libraries for advertisement. We use the dataset to investigates several aspects
of Android libraries, including their popularity and their proportion in
Android app code. Based on these datasets, we have further performed several
empirical investigations to confirm the motivations behind our work.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 10:49:27 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Li",
"Li",
""
],
[
"Bissyandé",
"Tegawendé F.",
""
],
[
"Klein",
"Jacques",
""
],
[
"Traon",
"Yves Le",
""
]
] | TITLE: An Investigation into the Use of Common Libraries in Android Apps
ABSTRACT: The packaging model of Android apps requires the entire code necessary for
the execution of an app to be shipped into one single apk file. Thus, an
analysis of Android apps often visits code which is not part of the
functionality delivered by the app. Such code is often contributed by the
common libraries which are used pervasively by all apps. Unfortunately, Android
analyses, e.g., for piggybacking detection and malware detection, can produce
inaccurate results if they do not take into account the case of library code,
which constitute noise in app features. Despite some efforts on investigating
Android libraries, the momentum of Android research has not yet produced a
complete set of common libraries to further support in-depth analysis of
Android apps. In this paper, we leverage a dataset of about 1.5 million apps
from Google Play to harvest potential common libraries, including advertisement
libraries. With several steps of refinements, we finally collect by far the
largest set of 1,113 libraries supporting common functionalities and 240
libraries for advertisement. We use the dataset to investigates several aspects
of Android libraries, including their popularity and their proportion in
Android app code. Based on these datasets, we have further performed several
empirical investigations to confirm the motivations behind our work.
| no_new_dataset | 0.784979 |
1511.06627 | Zhu Shizhan | Shizhan Zhu, Cheng Li, Chen Change Loy, Xiaoou Tang | Towards Arbitrary-View Face Alignment by Recommendation Trees | This is our original submission to ICCV 2015 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Learning to simultaneously handle face alignment of arbitrary views, e.g.
frontal and profile views, appears to be more challenging than we thought. The
difficulties lay in i) accommodating the complex appearance-shape relations
exhibited in different views, and ii) encompassing the varying landmark point
sets due to self-occlusion and different landmark protocols. Most existing
studies approach this problem via training multiple viewpoint-specific models,
and conduct head pose estimation for model selection. This solution is
intuitive but the performance is highly susceptible to inaccurate head pose
estimation. In this study, we address this shortcoming through learning an
Ensemble of Model Recommendation Trees (EMRT), which is capable of selecting
optimal model configuration without prior head pose estimation. The unified
framework seamlessly handles different viewpoints and landmark protocols, and
it is trained by optimising directly on landmark locations, thus yielding
superior results on arbitrary-view face alignment. This is the first study that
performs face alignment on the full AFLWdataset with faces of different views
including profile view. State-of-the-art performances are also reported on
MultiPIE and AFW datasets containing both frontaland profile-view faces.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 15:01:21 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Zhu",
"Shizhan",
""
],
[
"Li",
"Cheng",
""
],
[
"Loy",
"Chen Change",
""
],
[
"Tang",
"Xiaoou",
""
]
] | TITLE: Towards Arbitrary-View Face Alignment by Recommendation Trees
ABSTRACT: Learning to simultaneously handle face alignment of arbitrary views, e.g.
frontal and profile views, appears to be more challenging than we thought. The
difficulties lay in i) accommodating the complex appearance-shape relations
exhibited in different views, and ii) encompassing the varying landmark point
sets due to self-occlusion and different landmark protocols. Most existing
studies approach this problem via training multiple viewpoint-specific models,
and conduct head pose estimation for model selection. This solution is
intuitive but the performance is highly susceptible to inaccurate head pose
estimation. In this study, we address this shortcoming through learning an
Ensemble of Model Recommendation Trees (EMRT), which is capable of selecting
optimal model configuration without prior head pose estimation. The unified
framework seamlessly handles different viewpoints and landmark protocols, and
it is trained by optimising directly on landmark locations, thus yielding
superior results on arbitrary-view face alignment. This is the first study that
performs face alignment on the full AFLWdataset with faces of different views
including profile view. State-of-the-art performances are also reported on
MultiPIE and AFW datasets containing both frontaland profile-view faces.
| no_new_dataset | 0.946892 |
1511.06656 | Carlos Sarraute | Carlos Sarraute, Pablo Blanc, Javier Burroni | A Study of Age and Gender seen through Mobile Phone Usage Patterns in
Mexico | null | Proc. 2014 IEEE/ACM International Conference on Advances in Social
Networks Analysis and Mining (ASONAM), Beijing, China, 17-20 August 2014, pp.
836 - 843 | 10.1109/ASONAM.2014.6921683 | null | cs.SI physics.soc-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Mobile phone usage provides a wealth of information, which can be used to
better understand the demographic structure of a population. In this paper we
focus on the population of Mexican mobile phone users. Our first contribution
is an observational study of mobile phone usage according to gender and age
groups. We were able to detect significant differences in phone usage among
different subgroups of the population. Our second contribution is to provide a
novel methodology to predict demographic features (namely age and gender) of
unlabeled users by leveraging individual calling patterns, as well as the
structure of the communication graph. We provide details of the methodology and
show experimental results on a real world dataset that involves millions of
users.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 15:53:52 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Sarraute",
"Carlos",
""
],
[
"Blanc",
"Pablo",
""
],
[
"Burroni",
"Javier",
""
]
] | TITLE: A Study of Age and Gender seen through Mobile Phone Usage Patterns in
Mexico
ABSTRACT: Mobile phone usage provides a wealth of information, which can be used to
better understand the demographic structure of a population. In this paper we
focus on the population of Mexican mobile phone users. Our first contribution
is an observational study of mobile phone usage according to gender and age
groups. We were able to detect significant differences in phone usage among
different subgroups of the population. Our second contribution is to provide a
novel methodology to predict demographic features (namely age and gender) of
unlabeled users by leveraging individual calling patterns, as well as the
structure of the communication graph. We provide details of the methodology and
show experimental results on a real world dataset that involves millions of
users.
| no_new_dataset | 0.902007 |
1511.06663 | Cecilia Damon | Luca Talenti, Margaux Luck, Anastasia Yartseva, Nicolas Argy, Sandrine
Houz\'e and Cecilia Damon | L1 logistic regression as a feature selection step for training stable
classification trees for the prediction of severity criteria in imported
malaria | 18 pages, 10 figures, ICLR, computational science - Learning,
Imported Malaria, L1 logistic regression, Decision tree | null | null | null | cs.LG q-bio.QM stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multivariate classification methods using explanatory and predictive models
are necessary for characterizing subgroups of patients according to their risk
profiles. Popular methods include logistic regression and classification trees
with performances that vary according to the nature and the characteristics of
the dataset. In the context of imported malaria, we aimed at classifying
severity criteria based on a heterogeneous patient population. We investigated
these approaches by implementing two different strategies: L1 logistic
regression (L1LR) that models a single global solution and classification trees
that model multiple local solutions corresponding to discriminant subregions of
the feature space. For each strategy, we built a standard model, and a sparser
version of it. As an alternative to pruning, we explore a promising approach
that first constrains the tree model with an L1LR-based feature selection, an
approach we called L1LR-Tree. The objective is to decrease its vulnerability to
small data variations by removing variables corresponding to unstable local
phenomena. Our study is twofold: i) from a methodological perspective comparing
the performances and the stability of the three previous methods, i.e L1LR,
classification trees and L1LR-Tree, for the classification of severe forms of
imported malaria, and ii) from an applied perspective improving the actual
classification of severe forms of imported malaria by identifying more
personalized profiles predictive of several clinical criteria based on
variables dismissed for the clinical definition of the disease. The main
methodological results show that the combined method L1LR-Tree builds sparse
and stable models that significantly predicts the different severity criteria
and outperforms all the other methods in terms of accuracy.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 16:12:59 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Talenti",
"Luca",
""
],
[
"Luck",
"Margaux",
""
],
[
"Yartseva",
"Anastasia",
""
],
[
"Argy",
"Nicolas",
""
],
[
"Houzé",
"Sandrine",
""
],
[
"Damon",
"Cecilia",
""
]
] | TITLE: L1 logistic regression as a feature selection step for training stable
classification trees for the prediction of severity criteria in imported
malaria
ABSTRACT: Multivariate classification methods using explanatory and predictive models
are necessary for characterizing subgroups of patients according to their risk
profiles. Popular methods include logistic regression and classification trees
with performances that vary according to the nature and the characteristics of
the dataset. In the context of imported malaria, we aimed at classifying
severity criteria based on a heterogeneous patient population. We investigated
these approaches by implementing two different strategies: L1 logistic
regression (L1LR) that models a single global solution and classification trees
that model multiple local solutions corresponding to discriminant subregions of
the feature space. For each strategy, we built a standard model, and a sparser
version of it. As an alternative to pruning, we explore a promising approach
that first constrains the tree model with an L1LR-based feature selection, an
approach we called L1LR-Tree. The objective is to decrease its vulnerability to
small data variations by removing variables corresponding to unstable local
phenomena. Our study is twofold: i) from a methodological perspective comparing
the performances and the stability of the three previous methods, i.e L1LR,
classification trees and L1LR-Tree, for the classification of severe forms of
imported malaria, and ii) from an applied perspective improving the actual
classification of severe forms of imported malaria by identifying more
personalized profiles predictive of several clinical criteria based on
variables dismissed for the clinical definition of the disease. The main
methodological results show that the combined method L1LR-Tree builds sparse
and stable models that significantly predicts the different severity criteria
and outperforms all the other methods in terms of accuracy.
| no_new_dataset | 0.952618 |
1511.06674 | Marius Leordeanu | Anirudh Goyal and Marius Leordeanu | Stories in the Eye: Contextual Visual Interactions for Efficient Video
to Language Translation | null | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Integrating higher level visual and linguistic interpretations is at the
heart of human intelligence. As automatic visual category recognition in images
is approaching human performance, the high level understanding in the dynamic
spatiotemporal domain of videos and its translation into natural language is
still far from being solved. While most works on vision-to-text translations
use pre-learned or pre-established computational linguistic models, in this
paper we present an approach that uses vision alone to efficiently learn how to
translate into language the video content. We discover, in simple form, the
story played by main actors, while using only visual cues for representing
objects and their interactions. Our method learns in a hierarchical manner
higher level representations for recognizing subjects, actions and objects
involved, their relevant contextual background and their interaction to one
another over time. We have a three stage approach: first we take in
consideration features of the individual entities at the local level of
appearance, then we consider the relationship between these objects and actions
and their video background, and third, we consider their spatiotemporal
relations as inputs to classifiers at the highest level of interpretation.
Thus, our approach finds a coherent linguistic description of videos in the
form of a subject, verb and object based on their role played in the overall
visual story learned directly from training data, without using a known
language model. We test the efficiency of our approach on a large scale dataset
containing YouTube clips taken in the wild and demonstrate state-of-the-art
performance, often superior to current approaches that use more complex,
pre-learned linguistic knowledge.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 16:33:13 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Goyal",
"Anirudh",
""
],
[
"Leordeanu",
"Marius",
""
]
] | TITLE: Stories in the Eye: Contextual Visual Interactions for Efficient Video
to Language Translation
ABSTRACT: Integrating higher level visual and linguistic interpretations is at the
heart of human intelligence. As automatic visual category recognition in images
is approaching human performance, the high level understanding in the dynamic
spatiotemporal domain of videos and its translation into natural language is
still far from being solved. While most works on vision-to-text translations
use pre-learned or pre-established computational linguistic models, in this
paper we present an approach that uses vision alone to efficiently learn how to
translate into language the video content. We discover, in simple form, the
story played by main actors, while using only visual cues for representing
objects and their interactions. Our method learns in a hierarchical manner
higher level representations for recognizing subjects, actions and objects
involved, their relevant contextual background and their interaction to one
another over time. We have a three stage approach: first we take in
consideration features of the individual entities at the local level of
appearance, then we consider the relationship between these objects and actions
and their video background, and third, we consider their spatiotemporal
relations as inputs to classifiers at the highest level of interpretation.
Thus, our approach finds a coherent linguistic description of videos in the
form of a subject, verb and object based on their role played in the overall
visual story learned directly from training data, without using a known
language model. We test the efficiency of our approach on a large scale dataset
containing YouTube clips taken in the wild and demonstrate state-of-the-art
performance, often superior to current approaches that use more complex,
pre-learned linguistic knowledge.
| no_new_dataset | 0.947817 |
1511.06683 | Maksim Lapin | Maksim Lapin, Matthias Hein and Bernt Schiele | Top-k Multiclass SVM | NIPS 2015 | null | null | null | stat.ML cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Class ambiguity is typical in image classification problems with a large
number of classes. When classes are difficult to discriminate, it makes sense
to allow k guesses and evaluate classifiers based on the top-k error instead of
the standard zero-one loss. We propose top-k multiclass SVM as a direct method
to optimize for top-k performance. Our generalization of the well-known
multiclass SVM is based on a tight convex upper bound of the top-k error. We
propose a fast optimization scheme based on an efficient projection onto the
top-k simplex, which is of its own interest. Experiments on five datasets show
consistent improvements in top-k accuracy compared to various baselines.
| [
{
"version": "v1",
"created": "Fri, 20 Nov 2015 16:49:33 GMT"
}
] | 2015-11-23T00:00:00 | [
[
"Lapin",
"Maksim",
""
],
[
"Hein",
"Matthias",
""
],
[
"Schiele",
"Bernt",
""
]
] | TITLE: Top-k Multiclass SVM
ABSTRACT: Class ambiguity is typical in image classification problems with a large
number of classes. When classes are difficult to discriminate, it makes sense
to allow k guesses and evaluate classifiers based on the top-k error instead of
the standard zero-one loss. We propose top-k multiclass SVM as a direct method
to optimize for top-k performance. Our generalization of the well-known
multiclass SVM is based on a tight convex upper bound of the top-k error. We
propose a fast optimization scheme based on an efficient projection onto the
top-k simplex, which is of its own interest. Experiments on five datasets show
consistent improvements in top-k accuracy compared to various baselines.
| no_new_dataset | 0.954393 |
1409.7794 | Steven C.H. Hoi | Yue Wu, Steven C. H. Hoi, Tao Mei, Nenghai Yu | Large-scale Online Feature Selection for Ultra-high Dimensional Sparse
Data | 13 pages | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Feature selection with large-scale high-dimensional data is important yet
very challenging in machine learning and data mining. Online feature selection
is a promising new paradigm that is more efficient and scalable than batch
feature section methods, but the existing online approaches usually fall short
in their inferior efficacy as compared with batch approaches. In this paper, we
present a novel second-order online feature selection scheme that is simple yet
effective, very fast and extremely scalable to deal with large-scale ultra-high
dimensional sparse data streams. The basic idea is to improve the existing
first-order online feature selection methods by exploiting second-order
information for choosing the subset of important features with high confidence
weights. However, unlike many second-order learning methods that often suffer
from extra high computational cost, we devise a novel smart algorithm for
second-order online feature selection using a MaxHeap-based approach, which is
not only more effective than the existing first-order approaches, but also
significantly more efficient and scalable for large-scale feature selection
with ultra-high dimensional sparse data, as validated from our extensive
experiments. Impressively, on a billion-scale synthetic dataset (1-billion
dimensions, 1-billion nonzero features, and 1-million samples), our new
algorithm took only 8 minutes on a single PC, which is orders of magnitudes
faster than traditional batch approaches. \url{http://arxiv.org/abs/1409.7794}
| [
{
"version": "v1",
"created": "Sat, 27 Sep 2014 10:58:09 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Nov 2015 12:49:16 GMT"
},
{
"version": "v3",
"created": "Thu, 19 Nov 2015 14:21:21 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Wu",
"Yue",
""
],
[
"Hoi",
"Steven C. H.",
""
],
[
"Mei",
"Tao",
""
],
[
"Yu",
"Nenghai",
""
]
] | TITLE: Large-scale Online Feature Selection for Ultra-high Dimensional Sparse
Data
ABSTRACT: Feature selection with large-scale high-dimensional data is important yet
very challenging in machine learning and data mining. Online feature selection
is a promising new paradigm that is more efficient and scalable than batch
feature section methods, but the existing online approaches usually fall short
in their inferior efficacy as compared with batch approaches. In this paper, we
present a novel second-order online feature selection scheme that is simple yet
effective, very fast and extremely scalable to deal with large-scale ultra-high
dimensional sparse data streams. The basic idea is to improve the existing
first-order online feature selection methods by exploiting second-order
information for choosing the subset of important features with high confidence
weights. However, unlike many second-order learning methods that often suffer
from extra high computational cost, we devise a novel smart algorithm for
second-order online feature selection using a MaxHeap-based approach, which is
not only more effective than the existing first-order approaches, but also
significantly more efficient and scalable for large-scale feature selection
with ultra-high dimensional sparse data, as validated from our extensive
experiments. Impressively, on a billion-scale synthetic dataset (1-billion
dimensions, 1-billion nonzero features, and 1-million samples), our new
algorithm took only 8 minutes on a single PC, which is orders of magnitudes
faster than traditional batch approaches. \url{http://arxiv.org/abs/1409.7794}
| no_new_dataset | 0.950273 |
1506.03340 | Karl Moritz Hermann | Karl Moritz Hermann, Tom\'a\v{s} Ko\v{c}isk\'y, Edward Grefenstette,
Lasse Espeholt, Will Kay, Mustafa Suleyman and Phil Blunsom | Teaching Machines to Read and Comprehend | Appears in: Advances in Neural Information Processing Systems 28
(NIPS 2015). 14 pages, 13 figures | null | null | null | cs.CL cs.AI cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Teaching machines to read natural language documents remains an elusive
challenge. Machine reading systems can be tested on their ability to answer
questions posed on the contents of documents that they have seen, but until now
large scale training and test datasets have been missing for this type of
evaluation. In this work we define a new methodology that resolves this
bottleneck and provides large scale supervised reading comprehension data. This
allows us to develop a class of attention based deep neural networks that learn
to read real documents and answer complex questions with minimal prior
knowledge of language structure.
| [
{
"version": "v1",
"created": "Wed, 10 Jun 2015 14:54:39 GMT"
},
{
"version": "v2",
"created": "Thu, 1 Oct 2015 15:04:49 GMT"
},
{
"version": "v3",
"created": "Thu, 19 Nov 2015 15:43:23 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Hermann",
"Karl Moritz",
""
],
[
"Kočiský",
"Tomáš",
""
],
[
"Grefenstette",
"Edward",
""
],
[
"Espeholt",
"Lasse",
""
],
[
"Kay",
"Will",
""
],
[
"Suleyman",
"Mustafa",
""
],
[
"Blunsom",
"Phil",
""
]
] | TITLE: Teaching Machines to Read and Comprehend
ABSTRACT: Teaching machines to read natural language documents remains an elusive
challenge. Machine reading systems can be tested on their ability to answer
questions posed on the contents of documents that they have seen, but until now
large scale training and test datasets have been missing for this type of
evaluation. In this work we define a new methodology that resolves this
bottleneck and provides large scale supervised reading comprehension data. This
allows us to develop a class of attention based deep neural networks that learn
to read real documents and answer complex questions with minimal prior
knowledge of language structure.
| no_new_dataset | 0.950778 |
1507.02620 | Mircea Cimpoi | Mircea Cimpoi, Subhransu Maji, Iasonas Kokkinos, Andrea Vedaldi | Deep filter banks for texture recognition, description, and segmentation | 29 pages; 13 figures; 8 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visual textures have played a key role in image understanding because they
convey important semantics of images, and because texture representations that
pool local image descriptors in an orderless manner have had a tremendous
impact in diverse applications. In this paper we make several contributions to
texture understanding. First, instead of focusing on texture instance and
material category recognition, we propose a human-interpretable vocabulary of
texture attributes to describe common texture patterns, complemented by a new
describable texture dataset for benchmarking. Second, we look at the problem of
recognizing materials and texture attributes in realistic imaging conditions,
including when textures appear in clutter, developing corresponding benchmarks
on top of the recently proposed OpenSurfaces dataset. Third, we revisit classic
texture representations, including bag-of-visual-words and the Fisher vectors,
in the context of deep learning and show that these have excellent efficiency
and generalization properties if the convolutional layers of a deep model are
used as filter banks. We obtain in this manner state-of-the-art performance in
numerous datasets well beyond textures, an efficient method to apply deep
features to image regions, as well as benefit in transferring features from one
domain to another.
| [
{
"version": "v1",
"created": "Thu, 9 Jul 2015 17:55:30 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Nov 2015 23:10:52 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Cimpoi",
"Mircea",
""
],
[
"Maji",
"Subhransu",
""
],
[
"Kokkinos",
"Iasonas",
""
],
[
"Vedaldi",
"Andrea",
""
]
] | TITLE: Deep filter banks for texture recognition, description, and segmentation
ABSTRACT: Visual textures have played a key role in image understanding because they
convey important semantics of images, and because texture representations that
pool local image descriptors in an orderless manner have had a tremendous
impact in diverse applications. In this paper we make several contributions to
texture understanding. First, instead of focusing on texture instance and
material category recognition, we propose a human-interpretable vocabulary of
texture attributes to describe common texture patterns, complemented by a new
describable texture dataset for benchmarking. Second, we look at the problem of
recognizing materials and texture attributes in realistic imaging conditions,
including when textures appear in clutter, developing corresponding benchmarks
on top of the recently proposed OpenSurfaces dataset. Third, we revisit classic
texture representations, including bag-of-visual-words and the Fisher vectors,
in the context of deep learning and show that these have excellent efficiency
and generalization properties if the convolutional layers of a deep model are
used as filter banks. We obtain in this manner state-of-the-art performance in
numerous datasets well beyond textures, an efficient method to apply deep
features to image regions, as well as benefit in transferring features from one
domain to another.
| new_dataset | 0.964855 |
1510.06492 | Linus Hermansson | Linus Hermansson, Fredrik D. Johansson, and Osamu Watanabe | Generalized Shortest Path Kernel on Graphs | Short version presented at Discovery Science 2015 in Banff | null | null | null | cs.DS cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of classifying graphs using graph kernels. We define
a new graph kernel, called the generalized shortest path kernel, based on the
number and length of shortest paths between nodes. For our example
classification problem, we consider the task of classifying random graphs from
two well-known families, by the number of clusters they contain. We verify
empirically that the generalized shortest path kernel outperforms the original
shortest path kernel on a number of datasets. We give a theoretical analysis
for explaining our experimental results. In particular, we estimate
distributions of the expected feature vectors for the shortest path kernel and
the generalized shortest path kernel, and we show some evidence explaining why
our graph kernel outperforms the shortest path kernel for our graph
classification problem.
| [
{
"version": "v1",
"created": "Thu, 22 Oct 2015 05:49:31 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Hermansson",
"Linus",
""
],
[
"Johansson",
"Fredrik D.",
""
],
[
"Watanabe",
"Osamu",
""
]
] | TITLE: Generalized Shortest Path Kernel on Graphs
ABSTRACT: We consider the problem of classifying graphs using graph kernels. We define
a new graph kernel, called the generalized shortest path kernel, based on the
number and length of shortest paths between nodes. For our example
classification problem, we consider the task of classifying random graphs from
two well-known families, by the number of clusters they contain. We verify
empirically that the generalized shortest path kernel outperforms the original
shortest path kernel on a number of datasets. We give a theoretical analysis
for explaining our experimental results. In particular, we estimate
distributions of the expected feature vectors for the shortest path kernel and
the generalized shortest path kernel, and we show some evidence explaining why
our graph kernel outperforms the shortest path kernel for our graph
classification problem.
| no_new_dataset | 0.948251 |
1511.05788 | Michael Edwards | Michael Edwards and Jingjing Deng and Xianghua Xie | From Pose to Activity: Surveying Datasets and Introducing CONVERSE | Presentation of pose-based conversational human interaction dataset,
review of current appearance and depth based action recognition datasets,
public dataset, 38 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a review on the current state of publicly available datasets
within the human action recognition community; highlighting the revival of pose
based methods and recent progress of understanding person-person interaction
modeling. We categorize datasets regarding several key properties for usage as
a benchmark dataset; including the number of class labels, ground truths
provided, and application domain they occupy. We also consider the level of
abstraction of each dataset; grouping those that present actions, interactions
and higher level semantic activities. The survey identifies key appearance and
pose based datasets, noting a tendency for simplistic, emphasized, or scripted
action classes that are often readily definable by a stable collection of
sub-action gestures. There is a clear lack of datasets that provide closely
related actions, those that are not implicitly identified via a series of poses
and gestures, but rather a dynamic set of interactions. We therefore propose a
novel dataset that represents complex conversational interactions between two
individuals via 3D pose. 8 pairwise interactions describing 7 separate
conversation based scenarios were collected using two Kinect depth sensors. The
intention is to provide events that are constructed from numerous primitive
actions, interactions and motions, over a period of time; providing a set of
subtle action classes that are more representative of the real world, and a
challenge to currently developed recognition methodologies. We believe this is
among one of the first datasets devoted to conversational interaction
classification using 3D pose features and the attributed papers show this task
is indeed possible. The full dataset is made publicly available to the research
community at www.csvision.swansea.ac.uk/converse.
| [
{
"version": "v1",
"created": "Wed, 18 Nov 2015 14:03:55 GMT"
},
{
"version": "v2",
"created": "Thu, 19 Nov 2015 13:04:09 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Edwards",
"Michael",
""
],
[
"Deng",
"Jingjing",
""
],
[
"Xie",
"Xianghua",
""
]
] | TITLE: From Pose to Activity: Surveying Datasets and Introducing CONVERSE
ABSTRACT: We present a review on the current state of publicly available datasets
within the human action recognition community; highlighting the revival of pose
based methods and recent progress of understanding person-person interaction
modeling. We categorize datasets regarding several key properties for usage as
a benchmark dataset; including the number of class labels, ground truths
provided, and application domain they occupy. We also consider the level of
abstraction of each dataset; grouping those that present actions, interactions
and higher level semantic activities. The survey identifies key appearance and
pose based datasets, noting a tendency for simplistic, emphasized, or scripted
action classes that are often readily definable by a stable collection of
sub-action gestures. There is a clear lack of datasets that provide closely
related actions, those that are not implicitly identified via a series of poses
and gestures, but rather a dynamic set of interactions. We therefore propose a
novel dataset that represents complex conversational interactions between two
individuals via 3D pose. 8 pairwise interactions describing 7 separate
conversation based scenarios were collected using two Kinect depth sensors. The
intention is to provide events that are constructed from numerous primitive
actions, interactions and motions, over a period of time; providing a set of
subtle action classes that are more representative of the real world, and a
challenge to currently developed recognition methodologies. We believe this is
among one of the first datasets devoted to conversational interaction
classification using 3D pose features and the attributed papers show this task
is indeed possible. The full dataset is made publicly available to the research
community at www.csvision.swansea.ac.uk/converse.
| new_dataset | 0.965674 |
1511.06015 | Juan C. Caicedo | Juan C. Caicedo and Svetlana Lazebnik | Active Object Localization with Deep Reinforcement Learning | IEEE ICCV 2015 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present an active detection model for localizing objects in scenes. The
model is class-specific and allows an agent to focus attention on candidate
regions for identifying the correct location of a target object. This agent
learns to deform a bounding box using simple transformation actions, with the
goal of determining the most specific location of target objects following
top-down reasoning. The proposed localization agent is trained using deep
reinforcement learning, and evaluated on the Pascal VOC 2007 dataset. We show
that agents guided by the proposed model are able to localize a single instance
of an object after analyzing only between 11 and 25 regions in an image, and
obtain the best detection results among systems that do not use object
proposals for object localization.
| [
{
"version": "v1",
"created": "Wed, 18 Nov 2015 22:55:46 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Caicedo",
"Juan C.",
""
],
[
"Lazebnik",
"Svetlana",
""
]
] | TITLE: Active Object Localization with Deep Reinforcement Learning
ABSTRACT: We present an active detection model for localizing objects in scenes. The
model is class-specific and allows an agent to focus attention on candidate
regions for identifying the correct location of a target object. This agent
learns to deform a bounding box using simple transformation actions, with the
goal of determining the most specific location of target objects following
top-down reasoning. The proposed localization agent is trained using deep
reinforcement learning, and evaluated on the Pascal VOC 2007 dataset. We show
that agents guided by the proposed model are able to localize a single instance
of an object after analyzing only between 11 and 25 regions in an image, and
obtain the best detection results among systems that do not use object
proposals for object localization.
| no_new_dataset | 0.949529 |
1511.06072 | Sebastian Agethen | Sebastian Agethen, Winston H. Hsu | Mediated Experts for Deep Convolutional Networks | null | null | null | null | cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a new supervised architecture termed Mediated Mixture-of-Experts
(MMoE) that allows us to improve classification accuracy of Deep Convolutional
Networks (DCN). Our architecture achieves this with the help of expert
networks: A network is trained on a disjoint subset of a given dataset and then
run in parallel to other experts during deployment. A mediator is employed if
experts contradict each other. This allows our framework to naturally support
incremental learning, as adding new classes requires (re-)training of the new
expert only. We also propose two measures to control computational complexity:
An early-stopping mechanism halts experts that have low confidence in their
prediction. The system allows to trade-off accuracy and complexity without
further retraining. We also suggest to share low-level convolutional layers
between experts in an effort to avoid computation of a near-duplicate feature
set. We evaluate our system on a popular dataset and report improved accuracy
compared to a single model of same configuration.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 07:01:36 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Agethen",
"Sebastian",
""
],
[
"Hsu",
"Winston H.",
""
]
] | TITLE: Mediated Experts for Deep Convolutional Networks
ABSTRACT: We present a new supervised architecture termed Mediated Mixture-of-Experts
(MMoE) that allows us to improve classification accuracy of Deep Convolutional
Networks (DCN). Our architecture achieves this with the help of expert
networks: A network is trained on a disjoint subset of a given dataset and then
run in parallel to other experts during deployment. A mediator is employed if
experts contradict each other. This allows our framework to naturally support
incremental learning, as adding new classes requires (re-)training of the new
expert only. We also propose two measures to control computational complexity:
An early-stopping mechanism halts experts that have low confidence in their
prediction. The system allows to trade-off accuracy and complexity without
further retraining. We also suggest to share low-level convolutional layers
between experts in an effort to avoid computation of a near-duplicate feature
set. We evaluate our system on a popular dataset and report improved accuracy
compared to a single model of same configuration.
| no_new_dataset | 0.95018 |
1511.06147 | Abhimanyu Dubey | Abhimanyu Dubey, Nikhil Naik, Dan Raviv, Rahul Sukthankar and Ramesh
Raskar | Coreset-Based Adaptive Tracking | 8 pages, 5 figures, In submission to IEEE TPAMI (Transactions on
Pattern Analysis and Machine Intelligence) | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a method for learning from streaming visual data using a compact,
constant size representation of all the data that was seen until a given
moment. Specifically, we construct a 'coreset' representation of streaming data
using a parallelized algorithm, which is an approximation of a set with
relation to the squared distances between this set and all other points in its
ambient space. We learn an adaptive object appearance model from the coreset
tree in constant time and logarithmic space and use it for object tracking by
detection. Our method obtains excellent results for object tracking on three
standard datasets over more than 100 videos. The ability to summarize data
efficiently makes our method ideally suited for tracking in long videos in
presence of space and time constraints. We demonstrate this ability by
outperforming a variety of algorithms on the TLD dataset with 2685 frames on
average. This coreset based learning approach can be applied for both real-time
learning of small, varied data and fast learning of big data.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 12:59:20 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Dubey",
"Abhimanyu",
""
],
[
"Naik",
"Nikhil",
""
],
[
"Raviv",
"Dan",
""
],
[
"Sukthankar",
"Rahul",
""
],
[
"Raskar",
"Ramesh",
""
]
] | TITLE: Coreset-Based Adaptive Tracking
ABSTRACT: We propose a method for learning from streaming visual data using a compact,
constant size representation of all the data that was seen until a given
moment. Specifically, we construct a 'coreset' representation of streaming data
using a parallelized algorithm, which is an approximation of a set with
relation to the squared distances between this set and all other points in its
ambient space. We learn an adaptive object appearance model from the coreset
tree in constant time and logarithmic space and use it for object tracking by
detection. Our method obtains excellent results for object tracking on three
standard datasets over more than 100 videos. The ability to summarize data
efficiently makes our method ideally suited for tracking in long videos in
presence of space and time constraints. We demonstrate this ability by
outperforming a variety of algorithms on the TLD dataset with 2685 frames on
average. This coreset based learning approach can be applied for both real-time
learning of small, varied data and fast learning of big data.
| no_new_dataset | 0.94887 |
1511.06201 | Zhirong Wu | Zhirong Wu, Dahua Lin, Xiaoou Tang | Adjustable Bounded Rectifiers: Towards Deep Binary Representations | Under review as a conference paper at ICLR 2016 | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Binary representation is desirable for its memory efficiency, computation
speed and robustness. In this paper, we propose adjustable bounded rectifiers
to learn binary representations for deep neural networks. While hard
constraining representations across layers to be binary makes training
unreasonably difficult, we softly encourage activations to diverge from real
values to binary by approximating step functions. Our final representation is
completely binary. We test our approach on MNIST, CIFAR10, and ILSVRC2012
dataset, and systematically study the training dynamics of the binarization
process. Our approach can binarize the last layer representation without loss
of performance and binarize all the layers with reasonably small degradations.
The memory space that it saves may allow more sophisticated models to be
deployed, thus compensating the loss. To the best of our knowledge, this is the
first work to report results on current deep network architectures using
complete binary middle representations. Given the learned representations, we
find that the firing or inhibition of a binary neuron is usually associated
with a meaningful interpretation across different classes. This suggests that
the semantic structure of a neural network may be manifested through a guided
binarization process.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 15:14:02 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Wu",
"Zhirong",
""
],
[
"Lin",
"Dahua",
""
],
[
"Tang",
"Xiaoou",
""
]
] | TITLE: Adjustable Bounded Rectifiers: Towards Deep Binary Representations
ABSTRACT: Binary representation is desirable for its memory efficiency, computation
speed and robustness. In this paper, we propose adjustable bounded rectifiers
to learn binary representations for deep neural networks. While hard
constraining representations across layers to be binary makes training
unreasonably difficult, we softly encourage activations to diverge from real
values to binary by approximating step functions. Our final representation is
completely binary. We test our approach on MNIST, CIFAR10, and ILSVRC2012
dataset, and systematically study the training dynamics of the binarization
process. Our approach can binarize the last layer representation without loss
of performance and binarize all the layers with reasonably small degradations.
The memory space that it saves may allow more sophisticated models to be
deployed, thus compensating the loss. To the best of our knowledge, this is the
first work to report results on current deep network architectures using
complete binary middle representations. Given the learned representations, we
find that the firing or inhibition of a binary neuron is usually associated
with a meaningful interpretation across different classes. This suggests that
the semantic structure of a neural network may be manifested through a guided
binarization process.
| no_new_dataset | 0.947039 |
1511.06208 | Moshe Salhov | Moshe Salhov and Amit Bermanis and Guy Wolf and Amir Averbuch | Diffusion Representations | null | null | null | null | stat.ML cs.LG math.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion Maps framework is a kernel based method for manifold learning and
data analysis that defines diffusion similarities by imposing a Markovian
process on the given dataset. Analysis by this process uncovers the intrinsic
geometric structures in the data. Recently, it was suggested to replace the
standard kernel by a measure-based kernel that incorporates information about
the density of the data. Thus, the manifold assumption is replaced by a more
general measure-based assumption.
The measure-based diffusion kernel incorporates two separate independent
representations. The first determines a measure that correlates with a density
that represents normal behaviors and patterns in the data. The second consists
of the analyzed multidimensional data points.
In this paper, we present a representation framework for data analysis of
datasets that is based on a closed-form decomposition of the measure-based
kernel. The proposed representation preserves pairwise diffusion distances that
does not depend on the data size while being invariant to scale. For a
stationary data, no out-of-sample extension is needed for embedding newly
arrived data points in the representation space. Several aspects of the
presented methodology are demonstrated on analytically generated data.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 15:30:39 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Salhov",
"Moshe",
""
],
[
"Bermanis",
"Amit",
""
],
[
"Wolf",
"Guy",
""
],
[
"Averbuch",
"Amir",
""
]
] | TITLE: Diffusion Representations
ABSTRACT: Diffusion Maps framework is a kernel based method for manifold learning and
data analysis that defines diffusion similarities by imposing a Markovian
process on the given dataset. Analysis by this process uncovers the intrinsic
geometric structures in the data. Recently, it was suggested to replace the
standard kernel by a measure-based kernel that incorporates information about
the density of the data. Thus, the manifold assumption is replaced by a more
general measure-based assumption.
The measure-based diffusion kernel incorporates two separate independent
representations. The first determines a measure that correlates with a density
that represents normal behaviors and patterns in the data. The second consists
of the analyzed multidimensional data points.
In this paper, we present a representation framework for data analysis of
datasets that is based on a closed-form decomposition of the measure-based
kernel. The proposed representation preserves pairwise diffusion distances that
does not depend on the data size while being invariant to scale. For a
stationary data, no out-of-sample extension is needed for embedding newly
arrived data points in the representation space. Several aspects of the
presented methodology are demonstrated on analytically generated data.
| no_new_dataset | 0.948106 |
1511.06316 | Zinelabidine Boulkenafet Mr | Zinelabidine Boulkenafet, Jukka Komulainen, Abdenour Hadid | face anti-spoofing based on color texture analysis | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Research on face spoofing detection has mainly been focused on analyzing the
luminance of the face images, hence discarding the chrominance information
which can be useful for discriminating fake faces from genuine ones. In this
work, we propose a new face anti-spoofing method based on color texture
analysis. We analyze the joint color-texture information from the luminance and
the chrominance channels using a color local binary pattern descriptor. More
specifically, the feature histograms are extracted from each image band
separately. Extensive experiments on two benchmark datasets, namely CASIA face
anti-spoofing and Replay-Attack databases, showed excellent results compared to
the state-of-the-art. Most importantly, our inter-database evaluation depicts
that the proposed approach showed very promising generalization capabilities.
| [
{
"version": "v1",
"created": "Thu, 19 Nov 2015 19:28:20 GMT"
}
] | 2015-11-20T00:00:00 | [
[
"Boulkenafet",
"Zinelabidine",
""
],
[
"Komulainen",
"Jukka",
""
],
[
"Hadid",
"Abdenour",
""
]
] | TITLE: face anti-spoofing based on color texture analysis
ABSTRACT: Research on face spoofing detection has mainly been focused on analyzing the
luminance of the face images, hence discarding the chrominance information
which can be useful for discriminating fake faces from genuine ones. In this
work, we propose a new face anti-spoofing method based on color texture
analysis. We analyze the joint color-texture information from the luminance and
the chrominance channels using a color local binary pattern descriptor. More
specifically, the feature histograms are extracted from each image band
separately. Extensive experiments on two benchmark datasets, namely CASIA face
anti-spoofing and Replay-Attack databases, showed excellent results compared to
the state-of-the-art. Most importantly, our inter-database evaluation depicts
that the proposed approach showed very promising generalization capabilities.
| no_new_dataset | 0.948442 |
1506.03736 | Joseph Salmon | Eugene Ndiaye, Olivier Fercoq, Alexandre Gramfort, Joseph Salmon | GAP Safe screening rules for sparse multi-task and multi-class models | in Proceedings of the 29-th Conference on Neural Information
Processing Systems (NIPS), 2015 | null | null | null | stat.ML cs.LG math.OC stat.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High dimensional regression benefits from sparsity promoting regularizations.
Screening rules leverage the known sparsity of the solution by ignoring some
variables in the optimization, hence speeding up solvers. When the procedure is
proven not to discard features wrongly the rules are said to be \emph{safe}. In
this paper we derive new safe rules for generalized linear models regularized
with $\ell_1$ and $\ell_1/\ell_2$ norms. The rules are based on duality gap
computations and spherical safe regions whose diameters converge to zero. This
allows to discard safely more variables, in particular for low regularization
parameters. The GAP Safe rule can cope with any iterative solver and we
illustrate its performance on coordinate descent for multi-task Lasso, binary
and multinomial logistic regression, demonstrating significant speed ups on all
tested datasets with respect to previous safe rules.
| [
{
"version": "v1",
"created": "Thu, 11 Jun 2015 16:25:36 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Nov 2015 10:07:20 GMT"
}
] | 2015-11-19T00:00:00 | [
[
"Ndiaye",
"Eugene",
""
],
[
"Fercoq",
"Olivier",
""
],
[
"Gramfort",
"Alexandre",
""
],
[
"Salmon",
"Joseph",
""
]
] | TITLE: GAP Safe screening rules for sparse multi-task and multi-class models
ABSTRACT: High dimensional regression benefits from sparsity promoting regularizations.
Screening rules leverage the known sparsity of the solution by ignoring some
variables in the optimization, hence speeding up solvers. When the procedure is
proven not to discard features wrongly the rules are said to be \emph{safe}. In
this paper we derive new safe rules for generalized linear models regularized
with $\ell_1$ and $\ell_1/\ell_2$ norms. The rules are based on duality gap
computations and spherical safe regions whose diameters converge to zero. This
allows to discard safely more variables, in particular for low regularization
parameters. The GAP Safe rule can cope with any iterative solver and we
illustrate its performance on coordinate descent for multi-task Lasso, binary
and multinomial logistic regression, demonstrating significant speed ups on all
tested datasets with respect to previous safe rules.
| no_new_dataset | 0.945701 |
1506.04132 | Yingzhen Li | Yingzhen Li, Jose Miguel Hernandez-Lobato, Richard E. Turner | Stochastic Expectation Propagation | Published at NIPS 2015. 18 pages including supplementary | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Expectation propagation (EP) is a deterministic approximation algorithm that
is often used to perform approximate Bayesian parameter learning. EP
approximates the full intractable posterior distribution through a set of local
approximations that are iteratively refined for each datapoint. EP can offer
analytic and computational advantages over other approximations, such as
Variational Inference (VI), and is the method of choice for a number of models.
The local nature of EP appears to make it an ideal candidate for performing
Bayesian learning on large models in large-scale dataset settings. However, EP
has a crucial limitation in this context: the number of approximating factors
needs to increase with the number of data-points, N, which often entails a
prohibitively large memory overhead. This paper presents an extension to EP,
called stochastic expectation propagation (SEP), that maintains a global
posterior approximation (like VI) but updates it in a local way (like EP).
Experiments on a number of canonical learning problems using synthetic and
real-world datasets indicate that SEP performs almost as well as full EP, but
reduces the memory consumption by a factor of $N$. SEP is therefore ideally
suited to performing approximate Bayesian learning in the large model, large
dataset setting.
| [
{
"version": "v1",
"created": "Fri, 12 Jun 2015 19:51:06 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Nov 2015 10:52:17 GMT"
}
] | 2015-11-19T00:00:00 | [
[
"Li",
"Yingzhen",
""
],
[
"Hernandez-Lobato",
"Jose Miguel",
""
],
[
"Turner",
"Richard E.",
""
]
] | TITLE: Stochastic Expectation Propagation
ABSTRACT: Expectation propagation (EP) is a deterministic approximation algorithm that
is often used to perform approximate Bayesian parameter learning. EP
approximates the full intractable posterior distribution through a set of local
approximations that are iteratively refined for each datapoint. EP can offer
analytic and computational advantages over other approximations, such as
Variational Inference (VI), and is the method of choice for a number of models.
The local nature of EP appears to make it an ideal candidate for performing
Bayesian learning on large models in large-scale dataset settings. However, EP
has a crucial limitation in this context: the number of approximating factors
needs to increase with the number of data-points, N, which often entails a
prohibitively large memory overhead. This paper presents an extension to EP,
called stochastic expectation propagation (SEP), that maintains a global
posterior approximation (like VI) but updates it in a local way (like EP).
Experiments on a number of canonical learning problems using synthetic and
real-world datasets indicate that SEP performs almost as well as full EP, but
reduces the memory consumption by a factor of $N$. SEP is therefore ideally
suited to performing approximate Bayesian learning in the large model, large
dataset setting.
| no_new_dataset | 0.948489 |
1510.04709 | Desmond Elliott | Desmond Elliott, Stella Frank, Eva Hasler | Multilingual Image Description with Neural Sequence Models | Under review as a conference paper at ICLR 2016 | null | null | null | cs.CL cs.CV cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we present an approach to multi-language image description
bringing together insights from neural machine translation and neural image
description. To create a description of an image for a given target language,
our sequence generation models condition on feature vectors from the image, the
description from the source language, and/or a multimodal vector computed over
the image and a description in the source language. In image description
experiments on the IAPR-TC12 dataset of images aligned with English and German
sentences, we find significant and substantial improvements in BLEU4 and Meteor
scores for models trained over multiple languages, compared to a monolingual
baseline.
| [
{
"version": "v1",
"created": "Thu, 15 Oct 2015 20:29:21 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Nov 2015 17:04:35 GMT"
}
] | 2015-11-19T00:00:00 | [
[
"Elliott",
"Desmond",
""
],
[
"Frank",
"Stella",
""
],
[
"Hasler",
"Eva",
""
]
] | TITLE: Multilingual Image Description with Neural Sequence Models
ABSTRACT: In this paper we present an approach to multi-language image description
bringing together insights from neural machine translation and neural image
description. To create a description of an image for a given target language,
our sequence generation models condition on feature vectors from the image, the
description from the source language, and/or a multimodal vector computed over
the image and a description in the source language. In image description
experiments on the IAPR-TC12 dataset of images aligned with English and German
sentences, we find significant and substantial improvements in BLEU4 and Meteor
scores for models trained over multiple languages, compared to a monolingual
baseline.
| no_new_dataset | 0.945147 |
1511.04813 | Jing Lu | Jing Lu, Steven C.H. Hoi, Doyen Sahoo, Peilin Zhao | Budget Online Multiple Kernel Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Online learning with multiple kernels has gained increasing interests in
recent years and found many applications. For classification tasks, Online
Multiple Kernel Classification (OMKC), which learns a kernel based classifier
by seeking the optimal linear combination of a pool of single kernel
classifiers in an online fashion, achieves superior accuracy and enjoys great
flexibility compared with traditional single-kernel classifiers. Despite being
studied extensively, existing OMKC algorithms suffer from high computational
cost due to their unbounded numbers of support vectors. To overcome this
drawback, we present a novel framework of Budget Online Multiple Kernel
Learning (BOMKL) and propose a new Sparse Passive Aggressive learning to
perform effective budget online learning. Specifically, we adopt a simple yet
effective Bernoulli sampling to decide if an incoming instance should be added
to the current set of support vectors. By limiting the number of support
vectors, our method can significantly accelerate OMKC while maintaining
satisfactory accuracy that is comparable to that of the existing OMKC
algorithms. We theoretically prove that our new method achieves an optimal
regret bound in expectation, and empirically found that the proposed algorithm
outperforms various OMKC algorithms and can easily scale up to large-scale
datasets.
| [
{
"version": "v1",
"created": "Mon, 16 Nov 2015 03:40:50 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Nov 2015 08:08:43 GMT"
}
] | 2015-11-19T00:00:00 | [
[
"Lu",
"Jing",
""
],
[
"Hoi",
"Steven C. H.",
""
],
[
"Sahoo",
"Doyen",
""
],
[
"Zhao",
"Peilin",
""
]
] | TITLE: Budget Online Multiple Kernel Learning
ABSTRACT: Online learning with multiple kernels has gained increasing interests in
recent years and found many applications. For classification tasks, Online
Multiple Kernel Classification (OMKC), which learns a kernel based classifier
by seeking the optimal linear combination of a pool of single kernel
classifiers in an online fashion, achieves superior accuracy and enjoys great
flexibility compared with traditional single-kernel classifiers. Despite being
studied extensively, existing OMKC algorithms suffer from high computational
cost due to their unbounded numbers of support vectors. To overcome this
drawback, we present a novel framework of Budget Online Multiple Kernel
Learning (BOMKL) and propose a new Sparse Passive Aggressive learning to
perform effective budget online learning. Specifically, we adopt a simple yet
effective Bernoulli sampling to decide if an incoming instance should be added
to the current set of support vectors. By limiting the number of support
vectors, our method can significantly accelerate OMKC while maintaining
satisfactory accuracy that is comparable to that of the existing OMKC
algorithms. We theoretically prove that our new method achieves an optimal
regret bound in expectation, and empirically found that the proposed algorithm
outperforms various OMKC algorithms and can easily scale up to large-scale
datasets.
| no_new_dataset | 0.947332 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.