text
stringlengths 29
3.31k
| label
sequencelengths 1
11
|
|---|---|
The fully-convolutional network (FCN) with an encoder-decoder architecture
has been the standard paradigm for semantic segmentation. The encoder-decoder
architecture utilizes an encoder to capture multi-level feature maps, which are
incorporated into the final prediction by a decoder. As the context is crucial
for precise segmentation, tremendous effort has been made to extract such
information in an intelligent fashion, including employing dilated/atrous
convolutions or inserting attention modules. However, these endeavours are all
based on the FCN architecture with ResNet or other backbones, which cannot
fully exploit the context from the theoretical concept. By contrast, we propose
the Swin Transformer as the backbone to extract the context information and
design a novel decoder of densely connected feature aggregation module (DCFAM)
to restore the resolution and produce the segmentation map. The experimental
results on two remotely sensed semantic segmentation datasets demonstrate the
effectiveness of the proposed scheme. | [
"cs.CV"
] |
Face Super-Resolution (SR) is a subfield of the SR domain that specifically
targets the reconstruction of face images. The main challenge of face SR is to
restore essential facial features without distortion. We propose a novel face
SR method that generates photo-realistic 8x super-resolved face images with
fully retained facial details. To that end, we adopt a progressive training
method, which allows stable training by splitting the network into successive
steps, each producing output with a progressively higher resolution. We also
propose a novel facial attention loss and apply it at each step to focus on
restoring facial attributes in greater details by multiplying the pixel
difference and heatmap values. Lastly, we propose a compressed version of the
state-of-the-art face alignment network (FAN) for landmark heatmap extraction.
With the proposed FAN, we can extract the heatmaps suitable for face SR and
also reduce the overall training time. Experimental results verify that our
method outperforms state-of-the-art methods in both qualitative and
quantitative measurements, especially in perceptual quality. | [
"cs.CV"
] |
We propose a Convolutional Neural Network (CNN) based algorithm - StuffNet -
for object detection. In addition to the standard convolutional features
trained for region proposal and object detection [31], StuffNet uses
convolutional features trained for segmentation of objects and 'stuff'
(amorphous categories such as ground and water). Through experiments on Pascal
VOC 2010, we show the importance of features learnt from stuff segmentation for
improving object detection performance. StuffNet improves performance from
18.8% mAP to 23.9% mAP for small objects. We also devise a method to train
StuffNet on datasets that do not have stuff segmentation labels. Through
experiments on Pascal VOC 2007 and 2012, we demonstrate the effectiveness of
this method and show that StuffNet also significantly improves object detection
performance on such datasets. | [
"cs.CV"
] |
Multi-object tracking (MOT) in computer vision and cell tracking in
biomedical image analysis are two similar research fields, whose common aim is
to achieve instance level object detection/segmentation and associate such
objects across different video frames. However, one major difference between
these two tasks is that cell tracking also aim to detect mitosis (cell
division), which is typically not considered in MOT tasks. Therefore, the
acyclic oriented graphs matching (AOGM) has been used as de facto standard
evaluation metrics for cell tracking, rather than directly using the evaluation
metrics in computer vision, such as multiple object tracking accuracy (MOTA),
ID Switches (IDS), ID F1 Score (IDF1) etc. However, based on our experiments,
we realized that AOGM did not always function as expected for mitosis events.
In this paper, we exhibit the limitations of evaluating mitosis with AOGM using
both simulated and real cell tracking data. | [
"cs.CV",
"eess.IV",
"q-bio.QM"
] |
Relational regularized autoencoder (RAE) is a framework to learn the
distribution of data by minimizing a reconstruction loss together with a
relational regularization on the latent space. A recent attempt to reduce the
inner discrepancy between the prior and aggregated posterior distributions is
to incorporate sliced fused Gromov-Wasserstein (SFG) between these
distributions. That approach has a weakness since it treats every slicing
direction similarly, meanwhile several directions are not useful for the
discriminative task. To improve the discrepancy and consequently the relational
regularization, we propose a new relational discrepancy, named spherical sliced
fused Gromov Wasserstein (SSFG), that can find an important area of projections
characterized by a von Mises-Fisher distribution. Then, we introduce two
variants of SSFG to improve its performance. The first variant, named mixture
spherical sliced fused Gromov Wasserstein (MSSFG), replaces the vMF
distribution by a mixture of von Mises-Fisher distributions to capture multiple
important areas of directions that are far from each other. The second variant,
named power spherical sliced fused Gromov Wasserstein (PSSFG), replaces the vMF
distribution by a power spherical distribution to improve the sampling time in
high dimension settings. We then apply the new discrepancies to the RAE
framework to achieve its new variants. Finally, we conduct extensive
experiments to show that the new proposed autoencoders have favorable
performance in learning latent manifold structure, image generation, and
reconstruction. | [
"stat.ML",
"cs.LG"
] |
We present Generative Adversarial Capsule Network (CapsuleGAN), a framework
that uses capsule networks (CapsNets) instead of the standard convolutional
neural networks (CNNs) as discriminators within the generative adversarial
network (GAN) setting, while modeling image data. We provide guidelines for
designing CapsNet discriminators and the updated GAN objective function, which
incorporates the CapsNet margin loss, for training CapsuleGAN models. We show
that CapsuleGAN outperforms convolutional-GAN at modeling image data
distribution on MNIST and CIFAR-10 datasets, evaluated on the generative
adversarial metric and at semi-supervised image classification. | [
"stat.ML",
"cs.LG"
] |
Exploration is essential for solving complex Reinforcement Learning (RL)
tasks. Maximum State-Visitation Entropy (MSVE) formulates the exploration
problem as a well-defined policy optimization problem whose solution aims at
visiting all states as uniformly as possible. This is in contrast to standard
uncertainty-based approaches where exploration is transient and eventually
vanishes. However, existing approaches to MSVE are theoretically justified only
for discrete state-spaces as they are oblivious to the geometry of continuous
domains. We address this challenge by introducing Geometric Entropy
Maximisation (GEM), a new algorithm that maximises the geometry-aware Shannon
entropy of state-visits in both discrete and continuous domains. Our key
theoretical contribution is casting geometry-aware MSVE exploration as a
tractable problem of optimising a simple and novel noise-contrastive objective
function. In our experiments, we show the efficiency of GEM in solving several
RL problems with sparse rewards, compared against other deep RL exploration
approaches. | [
"cs.LG"
] |
One of the limiting factors in training data-driven, rare-event prediction
algorithms is the scarcity of the events of interest resulting in an extreme
imbalance in the data. There have been many methods introduced in the
literature for overcoming this issue; simple data manipulation through
undersampling and oversampling, utilizing cost-sensitive learning algorithms,
or by generating synthetic data points following the distribution of the
existing data. While synthetic data generation has recently received a great
deal of attention, there are real challenges involved in doing so for
high-dimensional data such as multivariate time series. In this study, we
explore the usefulness of the conditional generative adversarial network (CGAN)
as a means to perform data-informed oversampling in order to balance a large
dataset of multivariate time series. We utilize a flare forecasting benchmark
dataset, named SWAN-SF, and design two verification methods to both
quantitatively and qualitatively evaluate the similarity between the generated
minority and the ground-truth samples. We further assess the quality of the
generated samples by training a classical, supervised machine learning
algorithm on synthetic data, and testing the trained model on the unseen, real
data. The results show that the classifier trained on the data augmented with
the synthetic multivariate time series achieves a significant improvement
compared with the case where no augmentation is used. The popular flare
forecasting evaluation metrics, TSS and HSS, report 20-fold and 5-fold
improvements, respectively, indicating the remarkable statistical similarities,
and the usefulness of CGAN-based data generation for complicated tasks such as
flare forecasting. | [
"cs.LG"
] |
We discuss deep reinforcement learning in an overview style. We draw a big
picture, filled with details. We discuss six core elements, six important
mechanisms, and twelve applications, focusing on contemporary work, and in
historical contexts. We start with background of artificial intelligence,
machine learning, deep learning, and reinforcement learning (RL), with
resources. Next we discuss RL core elements, including value function, policy,
reward, model, exploration vs. exploitation, and representation. Then we
discuss important mechanisms for RL, including attention and memory,
unsupervised learning, hierarchical RL, multi-agent RL, relational RL, and
learning to learn. After that, we discuss RL applications, including games,
robotics, natural language processing (NLP), computer vision, finance, business
management, healthcare, education, energy, transportation, computer systems,
and, science, engineering, and art. Finally we summarize briefly, discuss
challenges and opportunities, and close with an epilogue. | [
"cs.LG",
"stat.ML"
] |
In this paper, we tackle the problem of egocentric action anticipation, i.e.,
predicting what actions the camera wearer will perform in the near future and
which objects they will interact with. Specifically, we contribute
Rolling-Unrolling LSTM, a learning architecture to anticipate actions from
egocentric videos. The method is based on three components: 1) an architecture
comprised of two LSTMs to model the sub-tasks of summarizing the past and
inferring the future, 2) a Sequence Completion Pre-Training technique which
encourages the LSTMs to focus on the different sub-tasks, and 3) a Modality
ATTention (MATT) mechanism to efficiently fuse multi-modal predictions
performed by processing RGB frames, optical flow fields and object-based
features. The proposed approach is validated on EPIC-Kitchens, EGTEA Gaze+ and
ActivityNet. The experiments show that the proposed architecture is
state-of-the-art in the domain of egocentric videos, achieving top performances
in the 2019 EPIC-Kitchens egocentric action anticipation challenge. The
approach also achieves competitive performance on ActivityNet with respect to
methods not based on unsupervised pre-training and generalizes to the tasks of
early action recognition and action recognition. To encourage research on this
challenging topic, we made our code, trained models, and pre-extracted features
available at our web page: http://iplab.dmi.unict.it/rulstm. | [
"cs.CV"
] |
LiDAR point-cloud segmentation is an important problem for many applications.
For large-scale point cloud segmentation, the \textit{de facto} method is to
project a 3D point cloud to get a 2D LiDAR image and use convolutions to
process it. Despite the similarity between regular RGB and LiDAR images, we
discover that the feature distribution of LiDAR images changes drastically at
different image locations. Using standard convolutions to process such LiDAR
images is problematic, as convolution filters pick up local features that are
only active in specific regions in the image. As a result, the capacity of the
network is under-utilized and the segmentation performance decreases. To fix
this, we propose Spatially-Adaptive Convolution (SAC) to adopt different
filters for different locations according to the input image. SAC can be
computed efficiently since it can be implemented as a series of element-wise
multiplications, im2col, and standard convolution. It is a general framework
such that several previous methods can be seen as special cases of SAC. Using
SAC, we build SqueezeSegV3 for LiDAR point-cloud segmentation and outperform
all previous published methods by at least 3.7% mIoU on the SemanticKITTI
benchmark with comparable inference speed. | [
"cs.CV"
] |
The principal contribution of this paper is a conceptual framework for
off-policy reinforcement learning, based on conditional expectations of
importance sampling ratios. This framework yields new perspectives and
understanding of existing off-policy algorithms, and reveals a broad space of
unexplored algorithms. We theoretically analyse this space, and concretely
investigate several algorithms that arise from this framework. | [
"cs.LG",
"stat.ML"
] |
A new method is proposed for removing text from natural images. The challenge
is to first accurately localize text on the stroke-level and then replace it
with a visually plausible background. Unlike previous methods that require
image patches to erase scene text, our method, namely ensconce network
(EnsNet), can operate end-to-end on a single image without any prior knowledge.
The overall structure is an end-to-end trainable FCN-ResNet-18 network with a
conditional generative adversarial network (cGAN). The feature of the former is
first enhanced by a novel lateral connection structure and then refined by four
carefully designed losses: multiscale regression loss and content loss, which
capture the global discrepancy of different level features; texture loss and
total variation loss, which primarily target filling the text region and
preserving the reality of the background. The latter is a novel local-sensitive
GAN, which attentively assesses the local consistency of the text erased
regions. Both qualitative and quantitative sensitivity experiments on synthetic
images and the ICDAR 2013 dataset demonstrate that each component of the EnsNet
is essential to achieve a good performance. Moreover, our EnsNet can
significantly outperform previous state-of-the-art methods in terms of all
metrics. In addition, a qualitative experiment conducted on the SMBNet dataset
further demonstrates that the proposed method can also preform well on general
object (such as pedestrians) removal tasks. EnsNet is extremely fast, which can
preform at 333 fps on an i5-8600 CPU device. | [
"cs.CV"
] |
Deep neural networks have shown excellent performance in stereo matching
task. Recently CNN-based methods have shown that stereo matching can be
formulated as a supervised learning task. However, less attention is paid on
the fusion of contextual semantic information and details. To tackle this
problem, we propose a network for disparity estimation based on abundant
contextual details and semantic information, called Multi-scale Features
Network (MSFNet). First, we design a new structure to encode rich semantic
information and fine-grained details by fusing multi-scale features. And we
combine the advantages of element-wise addition and concatenation, which is
conducive to merge semantic information with details. Second, a guidance
mechanism is introduced to guide the network to automatically focus more on the
unreliable regions. Third, we formulate the consistency check as an error map,
obtained by the low stage features with fine-grained details. Finally, we adopt
the consistency checking between the left feature and the synthetic left
feature to refine the initial disparity. Experiments on Scene Flow and KITTI
2015 benchmark demonstrated that the proposed method can achieve the
state-of-the-art performance. | [
"cs.CV"
] |
Recently there was an increasing interest in applications of graph neural
networks in non-Euclidean geometry; however, are non-Euclidean representations
always useful for graph learning tasks? For different problems such as node
classification and link prediction we compute hyperbolic embeddings and
conclude that for tasks that require global prediction consistency it might be
useful to use non-Euclidean embeddings, while for other tasks Euclidean models
are superior. To do so we first fix an issue of the existing models associated
with the optimization process at zero curvature. Current hyperbolic models deal
with gradients at the origin in ad-hoc manner, which is inefficient and can
lead to numerical instabilities. We solve the instabilities of
kappa-Stereographic model at zero curvature cases and evaluate the approach of
embedding graphs into the manifold in several graph representation learning
tasks. | [
"cs.LG",
"stat.ML"
] |
We propose RPSRNet - a novel end-to-end trainable deep neural network for
rigid point set registration. For this task, we use a novel $2^D$-tree
representation for the input point sets and a hierarchical deep feature
embedding in the neural network. An iterative transformation refinement module
in our network boosts the feature matching accuracy in the intermediate stages.
We achieve an inference speed of 12-15ms to register a pair of input point
clouds as large as 250K. Extensive evaluation on (i) KITTI LiDAR odometry and
(ii) ModelNet-40 datasets shows that our method outperforms prior
state-of-the-art methods - e.g., on the KITTI data set, DCP-v2 by1.3 and 1.5
times, and PointNetLK by 1.8 and 1.9 times better rotational and translational
accuracy respectively. Evaluation on ModelNet40 shows that RPSRNet is more
robust than other benchmark methods when the samples contain a significant
amount of noise and other disturbances. RPSRNet accurately registers point
clouds with non-uniform sampling densities, e.g., LiDAR data, which cannot be
processed by many existing deep-learning-based registration methods. | [
"cs.CV"
] |
We present a novel approach for image-animation of a source image by a
driving video, both depicting the same type of object. We do not assume the
existence of pose models and our method is able to animate arbitrary objects
without knowledge of the object's structure. Furthermore, both the driving
video and the source image are only seen during test-time. Our method is based
on a shared mask generator, which separates the foreground object from its
background, and captures the object's general pose and shape. A mask-refinement
module then replaces, in the mask extracted from the driver image, the identity
of the driver with the identity of the source. Conditioned on the source image,
the transformed mask is then decoded by a multi-scale generator that renders a
realistic image, in which the content of the source frame is animated by the
pose in the driving video. Due to lack of fully supervised data, we train on
the task of reconstructing frames from the same video the source image is taken
from. In order to control {the} source of the identity of the output frame, we
employ during training perturbations that remove the unwanted identity
information. Our method is shown to greatly outperform the state of the art
methods on multiple benchmarks. Our code and samples are available at
https://github.com/itsyoavshalev/Image-Animation-with-Perturbed-Masks | [
"cs.CV"
] |
The peaky behavior of CTC models is well known experimentally. However, an
understanding about why peaky behavior occurs is missing, and whether this is a
good property. We provide a formal analysis of the peaky behavior and gradient
descent convergence properties of the CTC loss and related training criteria.
Our analysis provides a deep understanding why peaky behavior occurs and when
it is suboptimal. On a simple example which should be trivial to learn for any
model, we prove that a feed-forward neural network trained with CTC from
uniform initialization converges towards peaky behavior with a 100% error rate.
Our analysis further explains why CTC only works well together with the blank
label. We further demonstrate that peaky behavior does not occur on other
related losses including a label prior model, and that this improves
convergence. | [
"cs.LG",
"cs.AI",
"cs.CL",
"cs.NE",
"cs.SD",
"eess.AS",
"math.ST",
"stat.TH"
] |
We consider the identifiability theory of probabilistic models and establish
sufficient conditions under which the representations learned by a very broad
family of conditional energy-based models are unique in function space, up to a
simple transformation. In our model family, the energy function is the
dot-product between two feature extractors, one for the dependent variable, and
one for the conditioning variable. We show that under mild conditions, the
features are unique up to scaling and permutation. Our results extend recent
developments in nonlinear ICA, and in fact, they lead to an important
generalization of ICA models. In particular, we show that our model can be used
for the estimation of the components in the framework of Independently
Modulated Component Analysis (IMCA), a new generalization of nonlinear ICA that
relaxes the independence assumption. A thorough empirical study shows that
representations learned by our model from real-world image datasets are
identifiable, and improve performance in transfer learning and semi-supervised
learning tasks. | [
"stat.ML",
"cs.LG"
] |
We study how stochastic differential equation (SDE) based ideas can inspire
new modifications to existing algorithms for a set of problems in computer
vision. Loosely speaking, our formulation is related to both explicit and
implicit strategies for data augmentation and group equivariance, but is
derived from new results in the SDE literature on estimating infinitesimal
generators of a class of stochastic processes. If and when there is nominal
agreement between the needs of an application/task and the inherent properties
and behavior of the types of processes that we can efficiently handle, we
obtain a very simple and efficient plug-in layer that can be incorporated
within any existing network architecture, with minimal modification and only a
few additional parameters. We show promising experiments on a number of vision
tasks including few shot learning, point cloud transformers and deep
variational segmentation obtaining efficiency or performance improvements. | [
"cs.CV",
"cs.LG"
] |
In hierarchical reinforcement learning a major challenge is determining
appropriate low-level policies. We propose an unsupervised learning scheme,
based on asymmetric self-play from Sukhbaatar et al. (2018), that automatically
learns a good representation of sub-goals in the environment and a low-level
policy that can execute them. A high-level policy can then direct the lower one
by generating a sequence of continuous sub-goal vectors. We evaluate our model
using Mazebase and Mujoco environments, including the challenging AntGather
task. Visualizations of the sub-goal embeddings reveal a logical decomposition
of tasks within the environment. Quantitatively, our approach obtains
compelling performance gains over non-hierarchical approaches. | [
"cs.LG",
"stat.ML"
] |
Modern applications of Bayesian inference involve models that are
sufficiently complex that the corresponding posterior distributions are
intractable and must be approximated. The most common approximation is based on
Markov chain Monte Carlo, but these can be expensive when the data set is large
and/or the model is complex, so more efficient variational approximations have
recently received considerable attention. The traditional variational methods,
that seek to minimize the Kullback--Leibler divergence between the posterior
and a relatively simple parametric family, provide accurate and efficient
estimation of the posterior mean, but often does not capture other moments, and
have limitations in terms of the models to which they can be applied. Here we
propose the construction of variational approximations based on minimizing the
Fisher divergence, and develop an efficient computational algorithm that can be
applied to a wide range of models without conjugacy or potentially unrealistic
mean-field assumptions. We demonstrate the superior performance of the proposed
method for the benchmark case of logistic regression. | [
"stat.ML",
"cs.LG",
"stat.CO",
"stat.ME"
] |
With the advances in capturing 2D or 3D skeleton data, skeleton-based action
recognition has received an increasing interest over the last years. As
skeleton data is commonly represented by graphs, graph convolutional networks
have been proposed for this task. While current graph convolutional networks
accurately recognize actions, they are too expensive for robotics applications
where limited computational resources are available. In this paper, we
therefore propose a highly efficient graph convolutional network that addresses
the limitations of previous works. This is achieved by a parallel structure
that gradually fuses motion and spatial information and by reducing the
temporal resolution as early as possible. Furthermore, we explicitly address
the issue that human poses can contain errors. To this end, the network first
refines the poses before they are further processed to recognize the action. We
therefore call the network Pose Refinement Graph Convolutional Network.
Compared to other graph convolutional networks, our network requires 86\%-93\%
less parameters and reduces the floating point operations by 89%-96% while
achieving a comparable accuracy. It therefore provides a much better trade-off
between accuracy, memory footprint and processing time, which makes it suitable
for robotics applications. | [
"cs.CV"
] |
MeanShift is a popular mode-seeking clustering algorithm used in a wide range
of applications in machine learning. However, it is known to be prohibitively
slow, with quadratic runtime per iteration. We propose MeanShift++, an
extremely fast mode-seeking algorithm based on MeanShift that uses a grid-based
approach to speed up the mean shift step, replacing the computationally
expensive neighbors search with a density-weighted mean of adjacent grid cells.
In addition, we show that this grid-based technique for density estimation
comes with theoretical guarantees. The runtime is linear in the number of
points and exponential in dimension, which makes MeanShift++ ideal on
low-dimensional applications such as image segmentation and object tracking. We
provide extensive experimental analysis showing that MeanShift++ can be more
than 10,000x faster than MeanShift with competitive clustering results on
benchmark datasets and nearly identical image segmentations as MeanShift.
Finally, we show promising results for object tracking. | [
"cs.CV",
"cs.LG"
] |
We design a simple reinforcement learning (RL) agent that implements an
optimistic version of $Q$-learning and establish through regret analysis that
this agent can operate with some level of competence in any environment. While
we leverage concepts from the literature on provably efficient RL, we consider
a general agent-environment interface and provide a novel agent design and
analysis. This level of generality positions our results to inform the design
of future agents for operation in complex real environments. We establish that,
as time progresses, our agent performs competitively relative to policies that
require longer times to evaluate. The time it takes to approach asymptotic
performance is polynomial in the complexity of the agent's state representation
and the time required to evaluate the best policy that the agent can represent.
Notably, there is no dependence on the complexity of the environment. The
ultimate per-period performance loss of the agent is bounded by a constant
multiple of a measure of distortion introduced by the agent's state
representation. This work is the first to establish that an algorithm
approaches this asymptotic condition within a tractable time frame. | [
"cs.LG",
"cs.AI"
] |
After building a classifier with modern tools of machine learning we
typically have a black box at hand that is able to predict well for unseen
data. Thus, we get an answer to the question what is the most likely label of a
given unseen data point. However, most methods will provide no answer why the
model predicted the particular label for a single instance and what features
were most influential for that particular instance. The only method that is
currently able to provide such explanations are decision trees. This paper
proposes a procedure which (based on a set of assumptions) allows to explain
the decisions of any classification method. | [
"stat.ML",
"cs.LG"
] |
A broad class of problems at the core of computational imaging, sensing, and
low-level computer vision reduces to the inverse problem of extracting latent
images that follow a prior distribution, from measurements taken under a known
physical image formation model. Traditionally, hand-crafted priors along with
iterative optimization methods have been used to solve such problems. In this
paper we present unrolled optimization with deep priors, a principled framework
for infusing knowledge of the image formation into deep networks that solve
inverse problems in imaging, inspired by classical iterative methods. We show
that instances of the framework outperform the state-of-the-art by a
substantial margin for a wide variety of imaging problems, such as denoising,
deblurring, and compressed sensing magnetic resonance imaging (MRI). Moreover,
we conduct experiments that explain how the framework is best used and why it
outperforms previous methods. | [
"cs.CV"
] |
Sample-efficient generalisation of reinforcement learning approaches have
always been a challenge, especially, for complex scenes with many components.
In this work, we introduce Plug and Play Markov Decision Processes, an
object-based representation that allows zero-shot integration of new objects
from known object classes. This is achieved by representing the global
transition dynamics as a union of local transition functions, each with respect
to one active object in the scene. Transition dynamics from an object class can
be pre-learnt and thus would be ready to use in a new environment. Each active
object is also endowed with its reward function. Since there is no central
reward function, addition or removal of objects can be handled efficiently by
only updating the reward functions of objects involved. A new transfer learning
mechanism is also proposed to adapt reward function in such cases. Experiments
show that our representation can achieve sample-efficiency in a variety of
set-ups. | [
"cs.LG"
] |
Geospatial object detection of remote sensing imagery has been attracting an
increasing interest in recent years, due to the rapid development in spaceborne
imaging. Most of previously proposed object detectors are very sensitive to
object deformations, such as scaling and rotation. To this end, we propose a
novel and efficient framework for geospatial object detection in this letter,
called Fourier-based rotation-invariant feature boosting (FRIFB). A
Fourier-based rotation-invariant feature is first generated in polar
coordinate. Then, the extracted features can be further structurally refined
using aggregate channel features. This leads to a faster feature computation
and more robust feature representation, which is good fitting for the coming
boosting learning. Finally, in the test phase, we achieve a fast pyramid
feature extraction by estimating a scale factor instead of directly collecting
all features from image pyramid. Extensive experiments are conducted on two
subsets of NWPU VHR-10 dataset, demonstrating the superiority and effectiveness
of the FRIFB compared to previous state-of-the-art methods. | [
"cs.CV"
] |
Single encoder-decoder methodologies for semantic segmentation are reaching
their peak in terms of segmentation quality and efficiency per number of
layers. To address these limitations, we propose a new architecture based on a
decoder which uses a set of shallow networks for capturing more information
content. The new decoder has a new topology of skip connections, namely
backward and stacked residual connections. In order to further improve the
architecture we introduce a weight function which aims to re-balance classes to
increase the attention of the networks to under-represented objects. We carried
out an extensive set of experiments that yielded state-of-the-art results for
the CamVid, Gatech and Freiburg Forest datasets. Moreover, to further prove the
effectiveness of our decoder, we conducted a set of experiments studying the
impact of our decoder to state-of-the-art segmentation techniques.
Additionally, we present a set of experiments augmenting semantic segmentation
with optical flow information, showing that motion clues can boost pure image
based semantic segmentation approaches. | [
"cs.CV",
"cs.RO"
] |
Rare diseases affect a relatively small number of people, which limits
investment in research for treatments and cures. Developing an efficient method
for rare disease detection is a crucial first step towards subsequent clinical
research. In this paper, we present a semi-supervised learning framework for
rare disease detection using generative adversarial networks. Our method takes
advantage of the large amount of unlabeled data for disease detection and
achieves the best results in terms of precision-recall score compared to
baseline techniques. | [
"cs.LG",
"stat.ML"
] |
This paper introduces the Deep Recurrent Attentive Writer (DRAW) neural
network architecture for image generation. DRAW networks combine a novel
spatial attention mechanism that mimics the foveation of the human eye, with a
sequential variational auto-encoding framework that allows for the iterative
construction of complex images. The system substantially improves on the state
of the art for generative models on MNIST, and, when trained on the Street View
House Numbers dataset, it generates images that cannot be distinguished from
real data with the naked eye. | [
"cs.CV",
"cs.LG",
"cs.NE"
] |
Deep learning models on graphs have achieved remarkable performance in
various graph analysis tasks, e.g., node classification, link prediction and
graph clustering. However, they expose uncertainty and unreliability against
the well-designed inputs, i.e., adversarial examples. Accordingly, a line of
studies have emerged for both attack and defense addressed in different graph
analysis tasks, leading to the arms race in graph adversarial learning.
Despite the booming works, there still lacks a unified problem definition and
a comprehensive review. To bridge this gap, we investigate and summarize the
existing works on graph adversarial learning tasks systemically. Specifically,
we survey and unify the existing works w.r.t. attack and defense in graph
analysis tasks, and give appropriate definitions and taxonomies at the same
time. Besides, we emphasize the importance of related evaluation metrics,
investigate and summarize them comprehensively. Hopefully, our works can
provide a comprehensive overview and offer insights for the relevant
researchers. More details of our works are available at
https://github.com/gitgiter/Graph-Adversarial-Learning. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Prior work presented the sentence tracker, a method for scoring how well a
sentence describes a video clip or alternatively how well a video clip depicts
a sentence. We present an improved method for optimizing the same cost function
employed by this prior work, reducing the space complexity from exponential in
the sentence length to polynomial, as well as producing a qualitatively
identical result in time polynomial in the sentence length instead of
exponential. Since this new method is plug-compatible with the prior method, it
can be used for the same applications: video retrieval with sentential queries,
generating sentential descriptions of video clips, and focusing the attention
of a tracker with a sentence, while allowing these applications to scale with
significantly larger numbers of object detections, word meanings modeled with
HMMs with significantly larger numbers of states, and significantly longer
sentences, with no appreciable degradation in quality of results. | [
"cs.CV"
] |
Face detection and recognition benchmarks have shifted toward more difficult
environments. The challenge presented in this paper addresses the next step in
the direction of automatic detection and identification of people from outdoor
surveillance cameras. While face detection has shown remarkable success in
images collected from the web, surveillance cameras include more diverse
occlusions, poses, weather conditions and image blur. Although face
verification or closed-set face identification have surpassed human
capabilities on some datasets, open-set identification is much more complex as
it needs to reject both unknown identities and false accepts from the face
detector. We show that unconstrained face detection can approach high detection
rates albeit with moderate false accept rates. By contrast, open-set face
recognition is currently weak and requires much more attention. | [
"cs.CV"
] |
Over the past few years, Generative Adversarial Networks (GANs) have garnered
increased interest among researchers in Computer Vision, with applications
including, but not limited to, image generation, translation, imputation, and
super-resolution. Nevertheless, no GAN-based method has been proposed in the
literature that can successfully represent, generate or translate 3D facial
shapes (meshes). This can be primarily attributed to two facts, namely that (a)
publicly available 3D face databases are scarce as well as limited in terms of
sample size and variability (e.g., few subjects, little diversity in race and
gender), and (b) mesh convolutions for deep networks present several challenges
that are not entirely tackled in the literature, leading to operator
approximations and model instability, often failing to preserve high-frequency
components of the distribution. As a result, linear methods such as Principal
Component Analysis (PCA) have been mainly utilized towards 3D shape analysis,
despite being unable to capture non-linearities and high frequency details of
the 3D face - such as eyelid and lip variations. In this work, we present
3DFaceGAN, the first GAN tailored towards modeling the distribution of 3D
facial surfaces, while retaining the high frequency details of 3D face shapes.
We conduct an extensive series of both qualitative and quantitative
experiments, where the merits of 3DFaceGAN are clearly demonstrated against
other, state-of-the-art methods in tasks such as 3D shape representation,
generation, and translation. | [
"cs.CV"
] |
We propose Cluster Pruning (CUP) for compressing and accelerating deep neural
networks. Our approach prunes similar filters by clustering them based on
features derived from both the incoming and outgoing weight connections. With
CUP, we overcome two limitations of prior work-(1) non-uniform pruning: CUP can
efficiently determine the ideal number of filters to prune in each layer of a
neural network. This is in contrast to prior methods that either prune all
layers uniformly or otherwise use resource-intensive methods such as manual
sensitivity analysis or reinforcement learning to determine the ideal number.
(2) Single-shot operation: We extend CUP to CUP-SS (for CUP single shot)
whereby pruning is integrated into the initial training phase itself. This
leads to large savings in training time compared to traditional pruning
pipelines. Through extensive evaluation on multiple datasets (MNIST, CIFAR-10,
and Imagenet) and models(VGG-16, Resnets-18/34/56) we show that CUP outperforms
recent state of the art. Specifically, CUP-SS achieves 2.2x flops reduction for
a Resnet-50 model trained on Imagenet while staying within 0.9% top-5 accuracy.
It saves over 14 hours in training time with respect to the original Resnet-50.
The code to reproduce results is available. | [
"cs.CV"
] |
We study the exploration problem in episodic MDPs with rich observations
generated from a small number of latent states. Under certain identifiability
assumptions, we demonstrate how to estimate a mapping from the observations to
latent states inductively through a sequence of regression and clustering steps
-- where previously decoded latent states provide labels for later regression
problems -- and use it to construct good exploration policies. We provide
finite-sample guarantees on the quality of the learned state decoding function
and exploration policies, and complement our theory with an empirical
evaluation on a class of hard exploration problems. Our method exponentially
improves over $Q$-learning with na\"ive exploration, even when $Q$-learning has
cheating access to latent states. | [
"cs.LG",
"stat.ML"
] |
This study compares the effectiveness and robustness of multi-class
categorization of Amazon product data using transfer learning on pre-trained
contextualized language models. Specifically, we fine-tuned BERT and XLNet, two
bidirectional models that have achieved state-of-the-art performance on many
natural language tasks and benchmarks, including text classification. While
existing classification studies and benchmarks focus on binary targets, with
the exception of ordinal ranking tasks, here we examine the robustness of such
models as the number of classes grows from 1 to 20. Our experiments demonstrate
an approximately linear decrease in performance metrics (i.e., precision,
recall, $F_1$ score, and accuracy) with the number of class labels. BERT
consistently outperforms XLNet using identical hyperparameters on the entire
range of class label quantities for categorizing products based on their
textual descriptions. BERT is also more affordable than XLNet in terms of the
computational cost (i.e., time and memory) required for training. In all cases
studied, the performance degradation rates were estimated to be 1% per
additional class label. | [
"stat.ML",
"cs.CL",
"cs.LG"
] |
The von Mises-Fisher (vMF) is a well-known density model for directional
random variables. The recent surge of the deep embedding methodologies for
high-dimensional structured data such as images or texts, aimed at extracting
salient directional information, can make the vMF model even more popular. In
this article, we will review the vMF model and its mixture, provide detailed
recipes of how to train the models, focusing on the maximum likelihood
estimators, in Python/PyTorch. In particular, implementation of vMF typically
suffers from the notorious numerical issue of the Bessel function evaluation in
the density normalizer, especially when the dimensionality is high, and we
address the issue using the MPMath library that supports arbitrary precision.
For the mixture learning, we provide both minibatch-based large-scale SGD
learning, as well as the EM algorithm which is a full batch estimator. For each
estimator/methodology, we test our implementation on some synthetic data, while
we also demonstrate the use case in a more realistic scenario of image
clustering. Our code is publicly available in
https://github.com/minyoungkim21/vmf-lib. | [
"cs.LG",
"cs.MS"
] |
Few-shot object detection has made substantial progressby representing novel
class objects using the feature representation learned upon a set of base class
objects. However,an implicit contradiction between novel class classification
and representation is unfortunately ignored. On the one hand, to achieve
accurate novel class classification, the distributions of either two base
classes must be far away fromeach other (max-margin). On the other hand, to
precisely represent novel classes, the distributions of base classes should be
close to each other to reduce the intra-class distance of novel classes
(min-margin). In this paper, we propose a class margin equilibrium (CME)
approach, with the aim to optimize both feature space partition and novel class
reconstruction in a systematic way. CME first converts the few-shot detection
problem to the few-shot classification problem by using a fully connected layer
to decouple localization features. CME then reserves adequate margin space for
novel classes by introducing simple-yet-effective class margin loss during
feature learning. Finally, CME pursues margin equilibrium by disturbing the
features of novel class instances in an adversarial min-max fashion.
Experiments on Pascal VOC and MS-COCO datasets show that CME significantly
improves upon two baseline detectors (up to $3\sim 5\%$ in average), achieving
state-of-the-art performance. Code is available at
https://github.com/Bohao-Lee/CME . | [
"cs.CV"
] |
Important information that relates to a specific topic in a document is often
organized in tabular format to assist readers with information retrieval and
comparison, which may be difficult to provide in natural language. However,
tabular data in unstructured digital documents, e.g., Portable Document Format
(PDF) and images, are difficult to parse into structured machine-readable
format, due to complexity and diversity in their structure and style. To
facilitate image-based table recognition with deep learning, we develop the
largest publicly available table recognition dataset PubTabNet
(https://github.com/ibm-aur-nlp/PubTabNet), containing 568k table images with
corresponding structured HTML representation. PubTabNet is automatically
generated by matching the XML and PDF representations of the scientific
articles in PubMed Central Open Access Subset (PMCOA). We also propose a novel
attention-based encoder-dual-decoder (EDD) architecture that converts images of
tables into HTML code. The model has a structure decoder which reconstructs the
table structure and helps the cell decoder to recognize cell content. In
addition, we propose a new Tree-Edit-Distance-based Similarity (TEDS) metric
for table recognition, which more appropriately captures multi-hop cell
misalignment and OCR errors than the pre-established metric. The experiments
demonstrate that the EDD model can accurately recognize complex tables solely
relying on the image representation, outperforming the state-of-the-art by 9.7%
absolute TEDS score. | [
"cs.CV"
] |
Automated capture of animal pose is transforming how we study neuroscience
and social behavior. Movements carry important social cues, but current methods
are not able to robustly estimate pose and shape of animals, particularly for
social animals such as birds, which are often occluded by each other and
objects in the environment. To address this problem, we first introduce a model
and multi-view optimization approach, which we use to capture the unique shape
and pose space displayed by live birds. We then introduce a pipeline and
experiments for keypoint, mask, pose, and shape regression that recovers
accurate avian postures from single views. Finally, we provide extensive
multi-view keypoint and mask annotations collected from a group of 15 social
birds housed together in an outdoor aviary. The project website with videos,
results, code, mesh model, and the Penn Aviary Dataset can be found at
https://marcbadger.github.io/avian-mesh. | [
"cs.CV",
"I.4.8"
] |
In this paper we study the matrix completion problem: Suppose $X \in {\mathbb
R}^{n_r \times n_c}$ is unknown except for a known upper bound $r$ on its rank.
By measuring a small number $m \ll n_r n_c$ of elements of $X$, is it possible
to recover $X$ exactly with noise-free measurements, or to construct a good
approximation of $X$ with noisy measurements? Existing solutions to these
problems involve sampling the elements uniformly and at random, and can
guarantee exact recovery of the unknown matrix only with high probability. In
this paper, we present a \textit{deterministic} sampling method for matrix
completion. We achieve this by choosing the sampling set as the edge set of an
asymmetric Ramanujan bigraph, and constrained nuclear norm minimization is the
recovery method. Specifically, we derive sufficient conditions under which the
unknown matrix is completed exactly with noise-free measurements, and is
approximately completed with noisy measurements, which we call "stable"
completion.
The conditions derived here are only sufficient and more restrictive than
random sampling. To study how close they are to being necessary, we conducted
numerical simulations on randomly generated low rank matrices, using the LPS
families of Ramanujan graphs. These simulations demonstrate two facts: (i) In
order to achieve exact completion, it appears sufficient to choose the degree
$d$ of the Ramanujan graph to be $\geq 3r$. (ii) There is a "phase transition,"
whereby the likelihood of success suddenly drops from 100\% to 0\% if the rank
is increased by just one or two beyond a critical value. The phase transition
phenomenon is well-known and well-studied in vector recovery using
$\ell_1$-norm minimization. However, it is less studied in matrix completion
and nuclear norm minimization, and not much understood. | [
"stat.ML",
"cs.LG",
"68T05"
] |
Patch-based stereo is nowadays a commonly used image-based technique for
dense 3D reconstruction in large scale multi-view applications. The typical
steps of such a pipeline can be summarized in stereo pair selection, depth map
computation, depth map refinement and, finally, fusion in order to generate a
complete and accurate representation of the scene in 3D. In this study, we aim
to support the standard dense 3D reconstruction of scenes as implemented in the
open source library OpenMVS by using semantic priors. To this end, during the
depth map fusion step, along with the depth consistency check between depth
maps of neighbouring views referring to the same part of the 3D scene, we
impose extra semantic constraints in order to remove possible errors and
selectively obtain segmented point clouds per label, boosting automation
towards this direction. I n order to reassure semantic coherence between
neighbouring views, additional semantic criterions can be considered, aiming to
elim inate mismatches of pixels belonging in different classes. | [
"cs.CV"
] |
Salient object detection has achieved great improvement by using the Fully
Convolution Network (FCN). However, the FCN-based U-shape architecture may
cause the dilution problem in the high-level semantic information during the
up-sample operations in the top-down pathway. Thus, it can weaken the ability
of salient object localization and produce degraded boundaries. To this end, in
order to overcome this limitation, we propose a novel pyramid self-attention
module (PSAM) and the adoption of an independent feature-complementing
strategy. In PSAM, self-attention layers are equipped after multi-scale pyramid
features to capture richer high-level features and bring larger receptive
fields to the model. In addition, a channel-wise attention module is also
employed to reduce the redundant features of the FPN and provide refined
results. Experimental analysis shows that the proposed PSAM effectively
contributes to the whole model so that it outperforms state-of-the-art results
over five challenging datasets. Finally, quantitative results show that PSAM
generates clear and integral salient maps which can provide further help to
other computer vision tasks, such as object detection and semantic
segmentation. | [
"cs.CV"
] |
With the goal of designing novel inhibitors for SARS-CoV-1 and SARS-CoV-2, we
propose the general molecule optimization framework, Molecular Neural Assay
Search (MONAS), consisting of three components: a property predictor which
identifies molecules with specific desirable properties, an energy model which
approximates the statistical similarity of a given molecule to known training
molecules, and a molecule search method. In this work, these components are
instantiated with graph neural networks (GNNs), Deep Energy Estimator Networks
(DEEN) and Monte Carlo tree search (MCTS), respectively. This implementation is
used to identify 120K molecules (out of 40-million explored) which the GNN
determined to be likely SARS-CoV-1 inhibitors, and, at the same time, are
statistically close to the dataset used to train the GNN. | [
"cs.LG",
"cs.AI",
"q-bio.QM"
] |
This thesis investigates unsupervised time series representation learning for
sequence prediction problems, i.e. generating nice-looking input samples given
a previous history, for high dimensional input sequences by decoupling the
static input representation from the recurrent sequence representation. We
introduce three models based on Generative Stochastic Networks (GSN) for
unsupervised sequence learning and prediction. Experimental results for these
three models are presented on pixels of sequential handwritten digit (MNIST)
data, videos of low-resolution bouncing balls, and motion capture data. The
main contribution of this thesis is to provide evidence that GSNs are a viable
framework to learn useful representations of complex sequential input data, and
to suggest a new framework for deep generative models to learn complex
sequences by decoupling static input representations from dynamic time
dependency representations. | [
"cs.LG",
"stat.ML"
] |
In this paper, we will investigate the contribution of color names for the
task of salient object detection. An input image is first converted to color
name space, which is consisted of 11 probabilistic channels. By exploiting a
surroundedness cue, we obtain a saliency map through a linear combination of a
set of sequential attention maps. To overcome the limitation of only using the
surroundedness cue, two global cues with respect to color names are invoked to
guide the computation of a weighted saliency map. Finally, we integrate the
above two saliency maps into a unified framework to generate the final result.
In addition, an improved post-processing procedure is introduced to effectively
suppress image backgrounds while uniformly highlight salient objects.
Experimental results show that the proposed model produces more accurate
saliency maps and performs well against twenty-one saliency models in terms of
three evaluation metrics on three public data sets. | [
"cs.CV",
"I.4"
] |
The goal of our paper is to semantically edit parts of an image matching a
given text that describes desired attributes (e.g., texture, colour, and
background), while preserving other contents that are irrelevant to the text.
To achieve this, we propose a novel generative adversarial network (ManiGAN),
which contains two key components: text-image affine combination module (ACM)
and detail correction module (DCM). The ACM selects image regions relevant to
the given text and then correlates the regions with corresponding semantic
words for effective manipulation. Meanwhile, it encodes original image features
to help reconstruct text-irrelevant contents. The DCM rectifies mismatched
attributes and completes missing contents of the synthetic image. Finally, we
suggest a new metric for evaluating image manipulation results, in terms of
both the generation of new attributes and the reconstruction of text-irrelevant
contents. Extensive experiments on the CUB and COCO datasets demonstrate the
superior performance of the proposed method. Code is available at
https://github.com/mrlibw/ManiGAN. | [
"cs.CV",
"cs.CL",
"cs.LG"
] |
Learning long-range behaviors on complex high-dimensional agents is a
fundamental problem in robot learning. For such tasks, we argue that
transferring learned information from a morphologically simpler agent can
massively improve the sample efficiency of a more complex one. To this end, we
propose a hierarchical decoupling of policies into two parts: an independently
learned low-level policy and a transferable high-level policy. To remedy poor
transfer performance due to mismatch in morphologies, we contribute two key
ideas. First, we show that incentivizing a complex agent's low-level to imitate
a simpler agent's low-level significantly improves zero-shot high-level
transfer. Second, we show that KL-regularized training of the high level
stabilizes learning and prevents mode-collapse. Finally, on a suite of publicly
released navigation and manipulation environments, we demonstrate the
applicability of hierarchical transfer on long-range tasks across morphologies.
Our code and videos can be found at
https://sites.google.com/berkeley.edu/morphology-transfer. | [
"cs.LG",
"cs.AI",
"cs.RO",
"stat.ML"
] |
Despite of the recent progress in agents that learn through interaction,
there are several challenges in terms of sample efficiency and generalization
across unseen behaviors during training. To mitigate these problems, we propose
and apply a first-order Meta-Learning algorithm called Bottom-Up Meta-Policy
Search (BUMPS), which works with two-phase optimization procedure: firstly, in
a meta-training phase, it distills few expert policies to create a meta-policy
capable of generalizing knowledge to unseen tasks during training; secondly, it
applies a fast adaptation strategy named Policy Filtering, which evaluates few
policies sampled from the meta-policy distribution and selects which best
solves the task. We conducted all experiments in the RoboCup 3D Soccer
Simulation domain, in the context of kick motion learning. We show that, given
our experimental setup, BUMPS works in scenarios where simple multi-task
Reinforcement Learning does not. Finally, we performed experiments in a way to
evaluate each component of the algorithm. | [
"cs.LG",
"cs.AI",
"cs.RO"
] |
Understanding and interpreting a 3d environment is a key challenge for
autonomous vehicles. Semantic segmentation of 3d point clouds combines 3d
information with semantics and thereby provides a valuable contribution to this
task. In many real-world applications, point clouds are generated by lidar
sensors in a consecutive fashion. Working with a time series instead of single
and independent frames enables the exploitation of temporal information. We
therefore propose a recurrent segmentation architecture (RNN), which takes a
single range image frame as input and exploits recursively aggregated temporal
information. An alignment strategy, which we call Temporal Memory Alignment,
uses ego motion to temporally align the memory between consecutive frames in
feature space. A Residual Network and ConvGRU are investigated for the memory
update. We demonstrate the benefits of the presented approach on two
large-scale datasets and compare it to several stateof-the-art methods. Our
approach ranks first on the SemanticKITTI multiple scan benchmark and achieves
state-of-the-art performance on the single scan benchmark. In addition, the
evaluation shows that the exploitation of temporal information significantly
improves segmentation results compared to a single frame approach. | [
"cs.CV"
] |
Conditional generative models have achieved considerable success in the past
few years, but usually require a lot of labeled data. Recently, ClusterGAN
combines GAN with an encoder to achieve remarkable clustering performance via
unsupervised conditional generation. However, it ignores the real conditional
distribution of data, which leads to generating less diverse samples for each
class and makes the encoder only achieve sub-optimal clustering performance.
Here, we propose a new unsupervised conditional generation framework, Double
Cycle-Consistent Conditional GAN (DC3-GAN), which can generate diverse
class-conditioned samples. We enforce the encoder and the generator of GAN to
form an encoder-generator pair in addition to the generator-encoder pair, which
enables us to avoid the low-diversity generation and the triviality of latent
features. We train the encoder-generator pair using real data, which can
indirectly estimate the real conditional distribution. Meanwhile, this
framework enforces the outputs of the encoder to match the inputs of GAN and
the prior noise distribution, which disentangles latent space into two parts:
one-hot discrete and continuous latent variables. The former can be directly
expressed as clusters and the latter represents remaining unspecified factors.
This work demonstrates that enhancing the diversity of unsupervised conditional
generated samples can improve the clustering performance. Experiments on
different benchmark datasets show that the proposed method outperforms existing
generative model-based clustering methods, and also achieves the optimal
disentanglement performance. | [
"cs.LG",
"cs.CV",
"stat.ML"
] |
We propose a new approach to image segmentation, which exploits the
advantages of both conditional random fields (CRFs) and decision trees. In the
literature, the potential functions of CRFs are mostly defined as a linear
combination of some pre-defined parametric models, and then methods like
structured support vector machines (SSVMs) are applied to learn those linear
coefficients. We instead formulate the unary and pairwise potentials as
nonparametric forests---ensembles of decision trees, and learn the ensemble
parameters and the trees in a unified optimization problem within the
large-margin framework. In this fashion, we easily achieve nonlinear learning
of potential functions on both unary and pairwise terms in CRFs. Moreover, we
learn class-wise decision trees for each object that appears in the image. Due
to the rich structure and flexibility of decision trees, our approach is
powerful in modelling complex data likelihoods and label relationships. The
resulting optimization problem is very challenging because it can have
exponentially many variables and constraints. We show that this challenging
optimization can be efficiently solved by combining a modified column
generation and cutting-planes techniques. Experimental results on both binary
(Graz-02, Weizmann horse, Oxford flower) and multi-class (MSRC-21, PASCAL VOC
2012) segmentation datasets demonstrate the power of the learned nonlinear
nonparametric potentials. | [
"cs.CV"
] |
The recent progress in multi-agent deep reinforcement learning(MADRL) makes
it more practical in real-world tasks, but its relatively poor scalability and
the partially observable constraints raise challenges to its performance and
deployment. Based on our intuitive observation that the human society could be
regarded as a large-scale partially observable environment, where each
individual has the function of communicating with neighbors and remembering its
own experience, we propose a novel network structure called hierarchical graph
recurrent network(HGRN) for multi-agent cooperation under partial
observability. Specifically, we construct the multi-agent system as a graph,
use the hierarchical graph attention network(HGAT) to achieve communication
between neighboring agents, and exploit GRU to enable agents to record
historical information. To encourage exploration and improve robustness, we
design a maximum-entropy learning method to learn stochastic policies of a
configurable target action entropy. Based on the above technologies, we
proposed a value-based MADRL algorithm called Soft-HGRN and its actor-critic
variant named SAC-HRGN. Experimental results based on three homogeneous tasks
and one heterogeneous environment not only show that our approach achieves
clear improvements compared with four baselines, but also demonstrates the
interpretability, scalability, and transferability of the proposed model.
Ablation studies prove the function and necessity of each component. | [
"cs.LG",
"cs.AI",
"cs.MA"
] |
Reinforcement learning (RL) enables robots to learn skills from interactions
with the real world. In practice, the unstructured step-based exploration used
in Deep RL -- often very successful in simulation -- leads to jerky motion
patterns on real robots. Consequences of the resulting shaky behavior are poor
exploration, or even damage to the robot. We address these issues by adapting
state-dependent exploration (SDE) to current Deep RL algorithms. To enable this
adaptation, we propose two extensions to the original SDE, using more general
features and re-sampling the noise periodically, which leads to a new
exploration method generalized state-dependent exploration (gSDE). We evaluate
gSDE both in simulation, on PyBullet continuous control tasks, and directly on
three different real robots: a tendon-driven elastic robot, a quadruped and an
RC car. The noise sampling interval of gSDE permits to have a compromise
between performance and smoothness, which allows training directly on the real
robots without loss of performance. The code is available at
https://github.com/DLR-RM/stable-baselines3. | [
"cs.LG",
"cs.RO",
"stat.ML"
] |
Convolutional Neural Networks (CNNs) are commonly thought to recognise
objects by learning increasingly complex representations of object shapes. Some
recent studies suggest a more important role of image textures. We here put
these conflicting hypotheses to a quantitative test by evaluating CNNs and
human observers on images with a texture-shape cue conflict. We show that
ImageNet-trained CNNs are strongly biased towards recognising textures rather
than shapes, which is in stark contrast to human behavioural evidence and
reveals fundamentally different classification strategies. We then demonstrate
that the same standard architecture (ResNet-50) that learns a texture-based
representation on ImageNet is able to learn a shape-based representation
instead when trained on "Stylized-ImageNet", a stylized version of ImageNet.
This provides a much better fit for human behavioural performance in our
well-controlled psychophysical lab setting (nine experiments totalling 48,560
psychophysical trials across 97 observers) and comes with a number of
unexpected emergent benefits such as improved object detection performance and
previously unseen robustness towards a wide range of image distortions,
highlighting advantages of a shape-based representation. | [
"cs.CV",
"cs.AI",
"cs.LG",
"q-bio.NC",
"stat.ML"
] |
Facial action unit (AU) recognition is a crucial task for facial expressions
analysis and has attracted extensive attention in the field of artificial
intelligence and computer vision. Existing works have either focused on
designing or learning complex regional feature representations, or delved into
various types of AU relationship modeling. Albeit with varying degrees of
progress, it is still arduous for existing methods to handle complex
situations. In this paper, we investigate how to integrate the semantic
relationship propagation between AUs in a deep neural network framework to
enhance the feature representation of facial regions, and propose an AU
semantic relationship embedded representation learning (SRERL) framework.
Specifically, by analyzing the symbiosis and mutual exclusion of AUs in various
facial expressions, we organize the facial AUs in the form of structured
knowledge-graph and integrate a Gated Graph Neural Network (GGNN) in a
multi-scale CNN framework to propagate node information through the graph for
generating enhanced AU representation. As the learned feature involves both the
appearance characteristics and the AU relationship reasoning, the proposed
model is more robust and can cope with more challenging cases, e.g.,
illumination change and partial occlusion. Extensive experiments on the two
public benchmarks demonstrate that our method outperforms the previous work and
achieves state of the art performance. | [
"cs.CV"
] |
Alzheimer's disease (AD) is the most prevalent form of dementia. Traditional
methods cannot achieve efficient and accurate diagnosis of AD. In this paper,
we introduce a novel method based on dynamic functional connectivity (dFC) that
can effectively capture changes in the brain. We compare and combine four
different types of features including amplitude of low-frequency fluctuation
(ALFF), regional homogeneity (ReHo), dFC and the adjacency matrix of different
brain structures between subjects. We use graph convolution network (GCN) which
consider the similarity of brain structure between patients to solve the
classification problem of non-Euclidean domains. The proposed method's accuracy
and the area under the receiver operating characteristic curve achieved 91.3%
and 98.4%. This result demonstrated that our proposed method can be used for
detecting AD. | [
"cs.CV",
"cs.LG",
"eess.IV"
] |
Sign Language Recognition (SLR) is a challenging research area in computer
vision. To tackle the annotation bottleneck in SLR, we formulate the problem of
Zero-Shot Sign Language Recognition (ZS-SLR) and propose a two-stream model
from two input modalities: RGB and Depth videos. To benefit from the vision
Transformer capabilities, we use two vision Transformer models, for human
detection and visual features representation. We configure a transformer
encoder-decoder architecture, as a fast and accurate human detection model, to
overcome the challenges of the current human detection models. Considering the
human keypoints, the detected human body is segmented into nine parts. A
spatio-temporal representation from human body is obtained using a vision
Transformer and a LSTM network. A semantic space maps the visual features to
the lingual embedding of the class labels via a Bidirectional Encoder
Representations from Transformers (BERT) model. We evaluated the proposed model
on four datasets, Montalbano II, MSR Daily Activity 3D, CAD-60, and NTU-60,
obtaining state-of-the-art results compared to state-of-the-art ZS-SLR models. | [
"cs.CV",
"cs.HC"
] |
Instance discriminative self-supervised representation learning has been
attracted attention thanks to its unsupervised nature and informative feature
representation for downstream tasks. In practice, it commonly uses a larger
number of negative samples than the number of supervised classes. However,
there is an inconsistency in the existing analysis; theoretically, a large
number of negative samples degrade classification performance on a downstream
supervised task, while empirically, they improve the performance. We provide a
novel framework to analyze this empirical result regarding negative samples
using the coupon collector's problem. Our bound can implicitly incorporate the
supervised loss of the downstream task in the self-supervised loss by
increasing the number of negative samples. We confirm that our proposed
analysis holds on real-world benchmark datasets. | [
"cs.LG",
"stat.ML"
] |
With the rapid development of spaceborne imaging techniques, object detection
in optical remote sensing imagery has drawn much attention in recent decades.
While many advanced works have been developed with powerful learning
algorithms, the incomplete feature representation still cannot meet the demand
for effectively and efficiently handling image deformations, particularly
objective scaling and rotation. To this end, we propose a novel object
detection framework, called optical remote sensing imagery detector (ORSIm
detector), integrating diverse channel features extraction, feature learning,
fast image pyramid matching, and boosting strategy. ORSIm detector adopts a
novel spatial-frequency channel feature (SFCF) by jointly considering the
rotation-invariant channel features constructed in frequency domain and the
original spatial channel features (e.g., color channel, gradient magnitude).
Subsequently, we refine SFCF using learning-based strategy in order to obtain
the high-level or semantically meaningful features. In the test phase, we
achieve a fast and coarsely-scaled channel computation by mathematically
estimating a scaling factor in the image domain. Extensive experimental results
conducted on the two different airborne datasets are performed to demonstrate
the superiority and effectiveness in comparison with previous state-of-the-art
methods. | [
"cs.CV"
] |
Omnidirectional images and spherical representations of $3D$ shapes cannot be
processed with conventional 2D convolutional neural networks (CNNs) as the
unwrapping leads to large distortion. Using fast implementations of spherical
and $SO(3)$ convolutions, researchers have recently developed deep learning
methods better suited for classifying spherical images. These newly proposed
convolutional layers naturally extend the notion of convolution to functions on
the unit sphere $S^2$ and the group of rotations $SO(3)$ and these layers are
equivariant to 3D rotations. In this paper, we consider the problem of
unsupervised learning of rotation-invariant representations for spherical
images. In particular, we carefully design an autoencoder architecture
consisting of $S^2$ and $SO(3)$ convolutional layers. As 3D rotations are often
a nuisance factor, the latent space is constrained to be exactly invariant to
these input transformations. As the rotation information is discarded in the
latent space, we craft a novel rotation-invariant loss function for training
the network. Extensive experiments on multiple datasets demonstrate the
usefulness of the learned representations on clustering, retrieval and
classification applications. | [
"cs.CV",
"cs.LG"
] |
To solve the problem that convolutional neural networks (CNNs) are difficult
to process non-grid type relational data like graphs, Kipf et al. proposed a
graph convolutional neural network (GCN). The core idea of the GCN is to
perform two-fold informational fusion for each node in a given graph during
each iteration: the fusion of graph structure information and the fusion of
node feature dimensions. Because of the characteristic of the combinatorial
generalizations, GCN has been widely used in the fields of scene semantic
relationship analysis, natural language processing and few-shot learning etc.
However, due to its two-fold informational fusion involves mathematical
irreversible calculations, it is hard to explain the decision reason for the
prediction of the each node classification. Unfortunately, most of the existing
attribution analysis methods concentrate on the models like CNNs, which are
utilized to process grid-like data. It is difficult to apply those analysis
methods to the GCN directly. It is because compared with the independence among
CNNs input data, there is correlation between the GCN input data. This
resulting in the existing attribution analysis methods can only obtain the
partial model contribution from the central node features to the final decision
of the GCN, but ignores the other model contribution from central node features
and its neighbor nodes features to that decision. To this end, we propose a
gradient attribution analysis method for the GCN called Node Attribution Method
(NAM), which can get the model contribution from not only the central node but
also its neighbor nodes to the GCN output. We also propose the Node Importance
Visualization (NIV) method to visualize the central node and its neighbor nodes
based on the value of the contribution... | [
"cs.LG"
] |
Omnidirectional lighting provides the foundation for achieving
spatially-variant photorealistic 3D rendering, a desirable property for mobile
augmented reality applications. However, in practice, estimating
omnidirectional lighting can be challenging due to limitations such as partial
panoramas of the rendering positions, and the inherent environment lighting and
mobile user dynamics. A new opportunity arises recently with the advancements
in mobile 3D vision, including built-in high-accuracy depth sensors and deep
learning-powered algorithms, which provide the means to better sense and
understand the physical surroundings. Centering the key idea of 3D vision, in
this work, we design an edge-assisted framework called Xihe to provide mobile
AR applications the ability to obtain accurate omnidirectional lighting
estimation in real time. Specifically, we develop a novel sampling technique
that efficiently compresses the raw point cloud input generated at the mobile
device. This technique is derived based on our empirical analysis of a recent
3D indoor dataset and plays a key role in our 3D vision-based lighting
estimator pipeline design. To achieve the real-time goal, we develop a tailored
GPU pipeline for on-device point cloud processing and use an encoding technique
that reduces network transmitted bytes. Finally, we present an adaptive
triggering strategy that allows Xihe to skip unnecessary lighting estimations
and a practical way to provide temporal coherent rendering integration with the
mobile AR ecosystem. We evaluate both the lighting estimation accuracy and time
of Xihe using a reference mobile application developed with Xihe's APIs. Our
results show that Xihe takes as fast as 20.67ms per lighting estimation and
achieves 9.4% better estimation accuracy than a state-of-the-art neural
network. | [
"cs.CV",
"cs.GR"
] |
Despite the tremendous success of deep neural networks in various learning
problems, it has been observed that adding an intentionally designed
adversarial perturbation to inputs of these architectures leads to erroneous
classification with high confidence in the prediction. In this work, we propose
a general framework based on the perturbation analysis of learning algorithms
which consists of convex programming and is able to recover many current
adversarial attacks as special cases. The framework can be used to propose
novel attacks against learning algorithms for classification and regression
tasks under various new constraints with closed form solutions in many
instances. In particular we derive new attacks against classification
algorithms which are shown to achieve comparable performances to notable
existing attacks. The framework is then used to generate adversarial
perturbations for regression tasks which include single pixel and single subset
attacks. By applying this method to autoencoding and image colorization tasks,
it is shown that adversarial perturbations can effectively perturb the output
of regression tasks as well. | [
"cs.LG",
"cs.AI",
"cs.IT",
"math.IT",
"stat.ML"
] |
Similar to humans and animals, deep artificial neural networks exhibit
critical periods during which a temporary stimulus deficit can impair the
development of a skill. The extent of the impairment depends on the onset and
length of the deficit window, as in animal models, and on the size of the
neural network. Deficits that do not affect low-level statistics, such as
vertical flipping of the images, have no lasting effect on performance and can
be overcome with further training. To better understand this phenomenon, we use
the Fisher Information of the weights to measure the effective connectivity
between layers of a network during training. Counterintuitively, information
rises rapidly in the early phases of training, and then decreases, preventing
redistribution of information resources in a phenomenon we refer to as a loss
of "Information Plasticity". Our analysis suggests that the first few epochs
are critical for the creation of strong connections that are optimal relative
to the input data distribution. Once such strong connections are created, they
do not appear to change during additional training. These findings suggest that
the initial learning transient, under-scrutinized compared to asymptotic
behavior, plays a key role in determining the outcome of the training process.
Our findings, combined with recent theoretical results in the literature, also
suggest that forgetting (decrease of information in the weights) is critical to
achieving invariance and disentanglement in representation learning. Finally,
critical periods are not restricted to biological systems, but can emerge
naturally in learning systems, whether biological or artificial, due to
fundamental constrains arising from learning dynamics and information
processing. | [
"cs.LG",
"q-bio.NC",
"stat.ML"
] |
Do you want to improve 1.0 AP for your object detector without any inference
cost and any change to your detector? Let us tell you such a recipe. It is
surprisingly simple: train your detector for an extra 12 epochs using cyclical
learning rates and then average these 12 checkpoints as your final detection
model}. This potent recipe is inspired by Stochastic Weights Averaging (SWA),
which is proposed in arXiv:1803.05407 for improving generalization in deep
neural networks. We found it also very effective in object detection. In this
technique report, we systematically investigate the effects of applying SWA to
object detection as well as instance segmentation. Through extensive
experiments, we discover the aforementioned workable policy of performing SWA
in object detection, and we consistently achieve $\sim$1.0 AP improvement over
various popular detectors on the challenging COCO benchmark, including Mask
RCNN, Faster RCNN, RetinaNet, FCOS, YOLOv3 and VFNet. We hope this work will
make more researchers in object detection know this technique and help them
train better object detectors. Code is available at:
https://github.com/hyz-xmaster/swa_object_detection . | [
"cs.CV"
] |
Given a sequence of sets, where each set contains an arbitrary number of
elements, the problem of temporal sets prediction aims to predict the elements
in the subsequent set. In practice, temporal sets prediction is much more
complex than predictive modelling of temporal events and time series, and is
still an open problem. Many possible existing methods, if adapted for the
problem of temporal sets prediction, usually follow a two-step strategy by
first projecting temporal sets into latent representations and then learning a
predictive model with the latent representations. The two-step approach often
leads to information loss and unsatisfactory prediction performance. In this
paper, we propose an integrated solution based on the deep neural networks for
temporal sets prediction. A unique perspective of our approach is to learn
element relationship by constructing set-level co-occurrence graph and then
perform graph convolutions on the dynamic relationship graphs. Moreover, we
design an attention-based module to adaptively learn the temporal dependency of
elements and sets. Finally, we provide a gated updating mechanism to find the
hidden shared patterns in different sequences and fuse both static and dynamic
information to improve the prediction performance. Experiments on real-world
data sets demonstrate that our approach can achieve competitive performances
even with a portion of the training data and can outperform existing methods
with a significant margin. | [
"cs.LG",
"cs.AI"
] |
With the proliferation of face image manipulation (FIM) techniques such as
Face2Face and Deepfake, more fake face images are spreading over the internet,
which brings serious challenges to public confidence. Face image forgery
detection has made considerable progresses in exposing specific FIM, but it is
still in scarcity of a robust fake face detector to expose face image forgeries
under complex scenarios such as with further compression, blurring, scaling,
etc. Due to the relatively fixed structure, convolutional neural network (CNN)
tends to learn image content representations. However, CNN should learn subtle
manipulation traces for image forensics tasks. Thus, we propose an adaptive
manipulation traces extraction network (AMTEN), which serves as pre-processing
to suppress image content and highlight manipulation traces. AMTEN exploits an
adaptive convolution layer to predict manipulation traces in the image, which
are reused in subsequent layers to maximize manipulation artifacts by updating
weights during the back-propagation pass. A fake face detector, namely
AMTENnet, is constructed by integrating AMTEN with CNN. Experimental results
prove that the proposed AMTEN achieves desirable pre-processing. When detecting
fake face images generated by various FIM techniques, AMTENnet achieves an
average accuracy up to 98.52%, which outperforms the state-of-the-art works.
When detecting face images with unknown post-processing operations, the
detector also achieves an average accuracy of 95.17%. | [
"cs.CV"
] |
In this paper, we consider a transfer Reinforcement Learning (RL) problem in
continuous state and action spaces, under unobserved contextual information.
For example, the context can represent the mental view of the world that an
expert agent has formed through past interactions with this world. We assume
that this context is not accessible to a learner agent who can only observe the
expert data. Then, our goal is to use the context-aware expert data to learn an
optimal context-unaware policy for the learner using only a few new data
samples. Such problems are typically solved using imitation learning that
assumes that both the expert and learner agents have access to the same
information. However, if the learner does not know the expert context, using
the expert data alone will result in a biased learner policy and will require
many new data samples to improve. To address this challenge, in this paper, we
formulate the learning problem as a causal bound-constrained Multi-Armed-Bandit
(MAB) problem. The arms of this MAB correspond to a set of basis policy
functions that can be initialized in an unsupervised way using the expert data
and represent the different expert behaviors affected by the unobserved
context. On the other hand, the MAB constraints correspond to causal bounds on
the accumulated rewards of these basis policy functions that we also compute
from the expert data. The solution to this MAB allows the learner agent to
select the best basis policy and improve it online. And the use of causal
bounds reduces the exploration variance and, therefore, improves the learning
rate. We provide numerical experiments on an autonomous driving example that
show that our proposed transfer RL method improves the learner's policy faster
compared to existing imitation learning methods and enjoys much lower variance
during training. | [
"cs.LG",
"cs.AI"
] |
3D image segmentation is one of the most important and ubiquitous problems in
medical image processing. It provides detailed quantitative analysis for
accurate disease diagnosis, abnormal detection, and classification. Currently
deep learning algorithms are widely used in medical image segmentation, most
algorithms trained models with full annotated datasets. However, obtaining
medical image datasets is very difficult and expensive, and full annotation of
3D medical image is a monotonous and time-consuming work. Partially labelling
informative slices in 3D images will be a great relief of manual annotation.
Sample selection strategies based on active learning have been proposed in the
field of 2D image, but few strategies focus on 3D images. In this paper, we
propose a sparse annotation strategy based on attention-guided active learning
for 3D medical image segmentation. Attention mechanism is used to improve
segmentation accuracy and estimate the segmentation accuracy of each slice. The
comparative experiments with three different strategies using datasets from the
developing human connectome project (dHCP) show that, our strategy only needs
15% to 20% annotated slices in brain extraction task and 30% to 35% annotated
slices in tissue segmentation task to achieve comparative results as full
annotation. | [
"cs.CV",
"eess.IV"
] |
Unsupervised learning of hierarchical representations has been one of the
most vibrant research directions in deep learning during recent years. In this
work we study biologically inspired unsupervised strategies in neural networks
based on local Hebbian learning. We propose new mechanisms to extend the
Bayesian Confidence Propagating Neural Network (BCPNN) architecture, and
demonstrate their capability for unsupervised learning of salient hidden
representations when tested on the MNIST dataset. | [
"cs.LG",
"cs.NE",
"stat.ML"
] |
A recently proposed methodology called the Horizontal Visibility Graph (HVG)
[Luque {\it et al.}, Phys. Rev. E., 80, 046103 (2009)] that constitutes a
geometrical simplification of the well known Visibility Graph algorithm [Lacasa
{\it et al.\/}, Proc. Natl. Sci. U.S.A. 105, 4972 (2008)], has been used to
study the distinction between deterministic and stochastic components in time
series [L. Lacasa and R. Toral, Phys. Rev. E., 82, 036120 (2010)].
Specifically, the authors propose that the node degree distribution of these
processes follows an exponential functional of the form $P(\kappa)\sim
\exp(-\lambda~\kappa)$, in which $\kappa$ is the node degree and $\lambda$ is a
positive parameter able to distinguish between deterministic (chaotic) and
stochastic (uncorrelated and correlated) dynamics. In this work, we investigate
the characteristics of the node degree distributions constructed by using HVG,
for time series corresponding to $28$ chaotic maps and $3$ different stochastic
processes. We thoroughly study the methodology proposed by Lacasa and Toral
finding several cases for which their hypothesis is not valid. We propose a
methodology that uses the HVG together with Information Theory quantifiers. An
extensive and careful analysis of the node degree distributions obtained by
applying HVG allow us to conclude that the Fisher-Shannon information plane is
a remarkable tool able to graphically represent the different nature,
deterministic or stochastic, of the systems under study. | [
"stat.ML",
"cs.IT",
"math.IT",
"nlin.CD"
] |
Digital image segmentation is the process of assigning distinct labels to
different objects in a digital image, and the fuzzy segmentation algorithm has
been successfully used in the segmentation of images from a wide variety of
sources. However, the traditional fuzzy segmentation algorithm fails to segment
objects that are characterized by textures whose patterns cannot be
successfully described by simple statistics computed over a very restricted
area. In this paper, we propose an extension of the fuzzy segmentation
algorithm that uses adaptive textural affinity functions to perform the
segmentation of such objects on bidimensional images. The adaptive affinity
functions compute their appropriate neighborhood size as they compute the
texture descriptors surrounding the seed spels (spatial elements), according to
the characteristics of the texture being processed. The algorithm then segments
the image with an appropriate neighborhood for each object. We performed
experiments on mosaic images that were composed using images from the Brodatz
database, and compared our results with the ones produced by a recently
published texture segmentation algorithm, showing the applicability of our
method. | [
"cs.CV",
"cs.AI",
"cs.GR"
] |
While deep learning has reshaped the classical motion capture pipeline,
generative, analysis-by-synthesis elements are still in use to recover fine
details if a high-quality 3D model of the user is available. Unfortunately,
obtaining such a model for every user a priori is challenging, time-consuming,
and limits the application scenarios. We propose a novel test-time optimization
approach for monocular motion capture that learns a volumetric body model of
the user in a self-supervised manner. To this end, our approach combines the
advantages of neural radiance fields with an articulated skeleton
representation. Our proposed skeleton embedding serves as a common reference
that links constraints across time, thereby reducing the number of required
camera views from traditionally dozens of calibrated cameras, down to a single
uncalibrated one. As a starting point, we employ the output of an off-the-shelf
model that predicts the 3D skeleton pose. The volumetric body shape and
appearance is then learned from scratch, while jointly refining the initial
pose estimate. Our approach is self-supervised and does not require any
additional ground truth labels for appearance, pose, or 3D shape. We
demonstrate that our novel combination of a discriminative pose estimation
technique with surface-free analysis-by-synthesis outperforms purely
discriminative monocular pose estimation approaches and generalizes well to
multiple views. | [
"cs.CV",
"cs.GR"
] |
Humans are capable of attributing latent mental contents such as beliefs or
intentions to others. The social skill is critical in daily life for reasoning
about the potential consequences of others' behaviors so as to plan ahead. It
is known that humans use such reasoning ability recursively by considering what
others believe about their own beliefs. In this paper, we start from level-$1$
recursion and introduce a probabilistic recursive reasoning (PR2) framework for
multi-agent reinforcement learning. Our hypothesis is that it is beneficial for
each agent to account for how the opponents would react to its future
behaviors. Under the PR2 framework, we adopt variational Bayes methods to
approximate the opponents' conditional policies, to which each agent finds the
best response and then improve their own policies. We develop
decentralized-training-decentralized-execution algorithms, namely PR2-Q and
PR2-Actor-Critic, that are proved to converge in the self-play scenarios when
there exists one Nash equilibrium. Our methods are tested on both the matrix
game and the differential game, which have a non-trivial equilibrium where
common gradient-based methods fail to converge. Our experiments show that it is
critical to reason about how the opponents believe about what the agent
believes. We expect our work to contribute a new idea of modeling the opponents
to the multi-agent reinforcement learning community. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Clustering is a fundamental unsupervised learning approach. Many clustering
algorithms -- such as $k$-means -- rely on the euclidean distance as a
similarity measure, which is often not the most relevant metric for high
dimensional data such as images. Learning a lower-dimensional embedding that
can better reflect the geometry of the dataset is therefore instrumental for
performance. We propose a new approach for this task where the embedding is
performed by a differentiable model such as a deep neural network. By rewriting
the $k$-means clustering algorithm as an optimal transport task, and adding an
entropic regularization, we derive a fully differentiable loss function that
can be minimized with respect to both the embedding parameters and the cluster
parameters via stochastic gradient descent. We show that this new formulation
generalizes a recently proposed state-of-the-art method based on soft-$k$-means
by adding constraints on the cluster sizes. Empirical evaluations on image
classification benchmarks suggest that compared to state-of-the-art methods,
our optimal transport-based approach provide better unsupervised accuracy and
does not require a pre-training phase. | [
"cs.LG",
"stat.ML"
] |
The policy gradient theorem is defined based on an objective with respect to
the initial distribution over states. In the discounted case, this results in
policies that are optimal for one distribution over initial states, but may not
be uniformly optimal for others, no matter where the agent starts from.
Furthermore, to obtain unbiased gradient estimates, the starting point of the
policy gradient estimator requires sampling states from a normalized discounted
weighting of states. However, the difficulty of estimating the normalized
discounted weighting of states, or the stationary state distribution, is quite
well-known. Additionally, the large sample complexity of policy gradient
methods is often attributed to insufficient exploration, and to remedy this, it
is often assumed that the restart distribution provides sufficient exploration
in these algorithms. In this work, we propose exploration in policy gradient
methods based on maximizing entropy of the discounted future state
distribution. The key contribution of our work includes providing a practically
feasible algorithm to estimate the normalized discounted weighting of states,
i.e, the \textit{discounted future state distribution}. We propose that
exploration can be achieved by entropy regularization with the discounted state
distribution in policy gradients, where a metric for maximal coverage of the
state space can be based on the entropy of the induced state distribution. The
proposed approach can be considered as a three time-scale algorithm and under
some mild technical conditions, we prove its convergence to a locally optimal
policy. Experimentally, we demonstrate usefulness of regularization with the
discounted future state distribution in terms of increased state space coverage
and faster learning on a range of complex tasks. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
A common challenge in reinforcement learning is how to convert the agent's
interactions with an environment into fast and robust learning. For instance,
earlier work makes use of domain knowledge to improve existing reinforcement
learning algorithms in complex tasks. While promising, previously acquired
knowledge is often costly and challenging to scale up. Instead, we decide to
consider problem knowledge with signals from quantities relevant to solve any
task, e.g., self-performance assessment and accurate expectations.
$\mathcal{V}^{ex}$ is such a quantity. It is the fraction of variance explained
by the value function $V$ and measures the discrepancy between $V$ and the
returns. Taking advantage of $\mathcal{V}^{ex}$, we propose MERL, a general
framework for structuring reinforcement learning by injecting problem knowledge
into policy gradient updates. As a result, the agent is not only optimized for
a reward but learns using problem-focused quantities provided by MERL,
applicable out-of-the-box to any task. In this paper: (a) We introduce and
define MERL, the multi-head reinforcement learning framework we use throughout
this work. (b) We conduct experiments across a variety of standard benchmark
environments, including 9 continuous control tasks, where results show improved
performance. (c) We demonstrate that MERL also improves transfer learning on a
set of challenging pixel-based tasks. (d) We ponder how MERL tackles the
problem of reward sparsity and better conditions the feature space of
reinforcement learning agents. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Koopman spectral analysis has attracted attention for nonlinear dynamical
systems since we can analyze nonlinear dynamics with a linear regime by
embedding data into a Koopman space by a nonlinear function. For the analysis,
we need to find appropriate embedding functions. Although several neural
network-based methods have been proposed for learning embedding functions,
existing methods require long time-series for training neural networks. This
limitation prohibits performing Koopman spectral analysis in applications where
only short time-series are available. In this paper, we propose a meta-learning
method for estimating embedding functions from unseen short time-series by
exploiting knowledge learned from related but different time-series. With the
proposed method, a representation of a given short time-series is obtained by a
bidirectional LSTM for extracting its properties. The embedding function of the
short time-series is modeled by a neural network that depends on the
time-series representation. By sharing the LSTM and neural networks across
multiple time-series, we can learn common knowledge from different time-series
while modeling time-series-specific embedding functions with the time-series
representation. Our model is trained such that the expected test prediction
error is minimized with the episodic training framework. We experimentally
demonstrate that the proposed method achieves better performance in terms of
eigenvalue estimation and future prediction than existing methods. | [
"stat.ML",
"cs.LG",
"math.DS"
] |
Entity interaction prediction is essential in many important applications
such as chemistry, biology, material science, and medical science. The problem
becomes quite challenging when each entity is represented by a complex
structure, namely structured entity, because two types of graphs are involved:
local graphs for structured entities and a global graph to capture the
interactions between structured entities. We observe that existing works on
structured entity interaction prediction cannot properly exploit the unique
graph of graphs model. In this paper, we propose a Graph of Graphs Neural
Network, namely GoGNN, which extracts the features in both structured entity
graphs and the entity interaction graph in a hierarchical way. We also propose
the dual-attention mechanism that enables the model to preserve the neighbor
importance in both levels of graphs. Extensive experiments on real-world
datasets show that GoGNN outperforms the state-of-the-art methods on two
representative structured entity interaction prediction tasks:
chemical-chemical interaction prediction and drug-drug interaction prediction.
Our code is available at Github. | [
"cs.LG",
"cs.SI"
] |
Guided depth super-resolution (GDSR) is a hot topic in multi-modal image
processing. The goal is to use high-resolution (HR) RGB images to provide extra
information on edges and object contours, so that low-resolution depth maps can
be upsampled to HR ones. To solve the issues of RGB texture over-transferred,
cross-modal feature extraction difficulty and unclear working mechanism of
modules in existing methods, we propose an advanced Discrete Cosine Transform
Network (DCTNet), which is composed of four components. Firstly, the paired
RGB/depth images are input into the semi-coupled feature extraction module. The
shared convolution kernels extract the cross-modal common features, and the
private kernels extract their unique features, respectively. Then the RGB
features are input into the edge attention mechanism to highlight the edges
useful for upsampling. Subsequently, in the Discrete Cosine Transform (DCT)
module, where DCT is employed to solve the optimization problem designed for
image domain GDSR. The solution is then extended to implement the multi-channel
RGB/depth features upsampling, which increases the rationality of DCTNet, and
is more flexible and effective than conventional methods. The final depth
prediction is output by the reconstruction module. Numerous qualitative and
quantitative experiments demonstrate the effectiveness of our method, which can
generate accurate and HR depth maps, surpassing state-of-the-art methods.
Meanwhile, the rationality of modules is also proved by ablation experiments. | [
"cs.CV"
] |
Conventional models for Visual Question Answering (VQA) explore deterministic
approaches with various types of image features, question features, and
attention mechanisms. However, there exist other modalities that can be
explored in addition to image and question pairs to bring extra information to
the models. In this work, we propose latent variable models for VQA where extra
information (e.g. captions and answer categories) are incorporated as latent
variables to improve inference, which in turn benefits question-answering
performance. Experiments on the VQA v2.0 benchmarking dataset demonstrate the
effectiveness of our proposed models in that they improve over strong
baselines, especially those that do not rely on extensive language-vision
pre-training. | [
"cs.CV",
"cs.AI",
"cs.CL"
] |
We present a versatile formulation of the convolution operation that we term
a "mapped convolution." The standard convolution operation implicitly samples
the pixel grid and computes a weighted sum. Our mapped convolution decouples
these two components, freeing the operation from the confines of the image grid
and allowing the kernel to process any type of structured data. As a test case,
we demonstrate its use by applying it to dense inference on spherical data. We
perform an in-depth study of existing spherical image convolution methods and
propose an improved sampling method for equirectangular images. Then, we
discuss the impact of data discretization when deriving a sampling function,
highlighting drawbacks of the cube map representation for spherical data.
Finally, we illustrate how mapped convolutions enable us to convolve directly
on a mesh by projecting the spherical image onto a geodesic grid and training
on the textured mesh. This method exceeds the state of the art for spherical
depth estimation by nearly 17%. Our findings suggest that mapped convolutions
can be instrumental in expanding the application scope of convolutional neural
networks. | [
"cs.CV"
] |
Deep Neural networks have gained lots of attention in recent years thanks to
the breakthroughs obtained in the field of Computer Vision. However, despite
their popularity, it has been shown that they provide limited robustness in
their predictions. In particular, it is possible to synthesise small
adversarial perturbations that imperceptibly modify a correctly classified
input data, making the network confidently misclassify it. This has led to a
plethora of different methods to try to improve robustness or detect the
presence of these perturbations. In this paper, we perform an analysis of
$\beta$-Variational Classifiers, a particular class of methods that not only
solve a specific classification task, but also provide a generative component
that is able to generate new samples from the input distribution. More in
details, we study their robustness and detection capabilities, together with
some novel insights on the generative part of the model. | [
"cs.LG",
"cs.CV",
"stat.ML"
] |
In this paper, we address the problem of makeup transfer, which aims at
transplanting the makeup from the reference face to the source face while
preserving the identity of the source. Existing makeup transfer methods have
made notable progress in generating realistic makeup faces, but do not perform
well in terms of color fidelity and spatial transformation. To tackle these
issues, we propose a novel Facial Attribute Transformer (FAT) and its variant
Spatial FAT for high-quality makeup transfer. Drawing inspirations from the
Transformer in NLP, FAT is able to model the semantic correspondences and
interactions between the source face and reference face, and then precisely
estimate and transfer the facial attributes. To further facilitate shape
deformation and transformation of facial parts, we also integrate thin plate
splines (TPS) into FAT, thus creating Spatial FAT, which is the first method
that can transfer geometric attributes in addition to color and texture.
Extensive qualitative and quantitative experiments demonstrate the
effectiveness and superiority of our proposed FATs in the following aspects:
(1) ensuring high-fidelity color transfer; (2) allowing for geometric
transformation of facial parts; (3) handling facial variations (such as poses
and shadows) and (4) supporting high-resolution face generation. | [
"cs.CV",
"cs.AI"
] |
Handwritten document-image binarization is a semantic segmentation process to
differentiate ink pixels from background pixels. It is one of the essential
steps towards character recognition, writer identification, and script-style
evolution analysis. The binarization task itself is challenging due to the vast
diversity of writing styles, inks, and paper materials. It is even more
difficult for historical manuscripts due to the aging and degradation of the
documents over time. One of such manuscripts is the Dead Sea Scrolls (DSS)
image collection, which poses extreme challenges for the existing binarization
techniques. This article proposes a new binarization technique for the DSS
images using the deep encoder-decoder networks. Although the artificial neural
network proposed here is primarily designed to binarize the DSS images, it can
be trained on different manuscript collections as well. Additionally, the use
of transfer learning makes the network already utilizable for a wide range of
handwritten documents, making it a unique multi-purpose tool for binarization.
Qualitative results and several quantitative comparisons using both historical
manuscripts and datasets from handwritten document image binarization
competition (H-DIBCO and DIBCO) exhibit the robustness and the effectiveness of
the system. The best performing network architecture proposed here is a variant
of the U-Net encoder-decoders. | [
"cs.CV"
] |
Although many methods have been proposed to deal with nature image
super-resolution (SR) and get impressive performance, the text images SR is not
good due to their ignorance of document images. In this paper, we propose a
matting-based dual generative adversarial network (mdGAN) for document image
SR. Firstly, the input image is decomposed into document text, foreground and
background layers using deep image matting. Then two parallel branches are
constructed to recover text boundary information and color information
respectively. Furthermore, in order to improve the restoration accuracy of
characters in output image, we use the input image's corresponding ground truth
text label as extra supervise information to refine the two-branch networks
during training. Experiments on real text images demonstrate that our method
outperforms several state-of-the-art methods quantitatively and qualitatively. | [
"cs.CV",
"cs.MM"
] |
In this paper, we propose a novel approach to solve the pose guided person
image generation task. We assume that the relation between pose and appearance
information can be described by a simple matrix operation in hidden space.
Based on this assumption, our method estimates a pose-invariant feature matrix
for each identity, and uses it to predict the target appearance conditioned on
the target pose. The estimation process is formulated as a p-norm regression
problem in hidden space. By utilizing the differentiation of the solution of
this regression problem, the parameters of the whole framework can be trained
in an end-to-end manner. While most previous works are only applicable to the
supervised training and single-shot generation scenario, our method can be
easily adapted to unsupervised training and multi-shot generation. Extensive
experiments on the challenging Market-1501 dataset show that our method yields
competitive performance in all the aforementioned variant scenarios. | [
"cs.CV"
] |
Let $A$ be a matrix with its pseudo-matrix $A^{\dagger}$ and set
$S=I-A^{\dagger}A$. We prove that, after re-ordering the columns of $A$, the
matrix $S$ has a block-diagonal form where each block corresponds to a set of
linearly dependent columns. This allows us to identify redundant columns in
$A$. We explore some applications in supervised and unsupervised learning,
specially feature selection, clustering, and sensitivity of solutions of least
squares solutions. | [
"cs.LG",
"stat.ML"
] |
Distortion quantification of point clouds plays a stealth, yet vital role in
a wide range of human and machine perception tasks. For human perception tasks,
a distortion quantification can substitute subjective experiments to guide 3D
visualization; while for machine perception tasks, a distortion quantification
can work as a loss function to guide the training of deep neural networks for
unsupervised learning tasks. To handle a variety of demands in many
applications, a distortion quantification needs to be distortion discriminable,
differentiable, and have a low computational complexity. Currently, however,
there is a lack of a general distortion quantification that can satisfy all
three conditions. To fill this gap, this work proposes multiscale potential
energy discrepancy (MPED), a distortion quantification to measure point cloud
geometry and color difference. By evaluating at various neighborhood sizes, the
proposed MPED achieves global-local tradeoffs, capturing distortion in a
multiscale fashion. Extensive experimental studies validate MPED's superiority
for both human and machine perception tasks. | [
"cs.CV",
"eess.IV"
] |
The goal of few-shot classification is to classify new categories with few
labeled examples within each class. Nowadays, the excellent performance in
handling few-shot classification problems is shown by metric-based
meta-learning methods. However, it is very hard for previous methods to
discriminate the fine-grained sub-categories in the embedding space without
fine-grained labels. This may lead to unsatisfactory generalization to
fine-grained subcategories, and thus affects model interpretation. To tackle
this problem, we introduce the contrastive loss into few-shot classification
for learning latent fine-grained structure in the embedding space. Furthermore,
to overcome the drawbacks of random image transformation used in current
contrastive learning in producing noisy and inaccurate image pairs (i.e.,
views), we develop a learning-to-learn algorithm to automatically generate
different views of the same image. Extensive experiments on standard few-shot
learning benchmarks demonstrate the superiority of our method. | [
"cs.CV"
] |
Precise financial series predicting has long been a difficult problem because
of unstableness and many noises within the series. Although Traditional time
series models like ARIMA and GARCH have been researched and proved to be
effective in predicting, their performances are still far from satisfying.
Machine Learning, as an emerging research field in recent years, has brought
about many incredible improvements in tasks such as regressing and classifying,
and it's also promising to exploit the methodology in financial time series
predicting. In this paper, the predicting precision of financial time series
between traditional time series models and mainstream machine learning models
including some state-of-the-art ones of deep learning are compared through
experiment using real stock index data from history. The result shows that
machine learning as a modern method far surpasses traditional models in
precision. | [
"cs.LG",
"q-fin.ST"
] |
Message-passing has proved to be an effective way to design graph neural
networks, as it is able to leverage both permutation equivariance and an
inductive bias towards learning local structures in order to achieve good
generalization. However, current message-passing architectures have a limited
representation power and fail to learn basic topological properties of graphs.
We address this problem and propose a powerful and equivariant message-passing
framework based on two ideas: first, we propagate a one-hot encoding of the
nodes, in addition to the features, in order to learn a local context matrix
around each node. This matrix contains rich local information about both
features and topology and can eventually be pooled to build node
representations. Second, we propose methods for the parametrization of the
message and update functions that ensure permutation equivariance. Having a
representation that is independent of the specific choice of the one-hot
encoding permits inductive reasoning and leads to better generalization
properties. Experimentally, our model can predict various graph topological
properties on synthetic data more accurately than previous methods and achieves
state-of-the-art results on molecular graph regression on the ZINC dataset. | [
"cs.LG",
"stat.ML"
] |
We show in this note that the Sobolev Discrepancy introduced in Mroueh et al
in the context of generative adversarial networks, is actually the weighted
negative Sobolev norm $||.||_{\dot{H}^{-1}(\nu_q)}$, that is known to linearize
the Wasserstein $W_2$ distance and plays a fundamental role in the dynamic
formulation of optimal transport of Benamou and Brenier. Given a Kernel with
finite dimensional feature map we show that the Sobolev discrepancy can be
approximated from finite samples. Assuming this discrepancy is finite, the
error depends on the approximation error in the function space induced by the
finite dimensional feature space kernel and on a statistical error due to the
finite sample approximation. | [
"cs.LG",
"stat.ML"
] |
The past decade has seen a rapid penetration of electric vehicles (EV) in the
market, more and more logistics and transportation companies start to deploy
EVs for service provision. In order to model the operations of a commercial EV
fleet, we utilize the EV routing problem with time windows (EVRPTW). In this
research, we propose an end-to-end deep reinforcement learning framework to
solve the EVRPTW. In particular, we develop an attention model incorporating
the pointer network and a graph embedding technique to parameterize a
stochastic policy for solving the EVRPTW. The model is then trained using
policy gradient with rollout baseline. Our numerical studies show that the
proposed model is able to efficiently solve EVRPTW instances of large sizes
that are not solvable with any existing approaches. | [
"cs.LG",
"cs.AI",
"math.OC",
"stat.ML"
] |
Explainable Artificial Intelligence (XAI), i.e., the development of more
transparent and interpretable AI models, has gained increased traction over the
last few years. This is due to the fact that, in conjunction with their growth
into powerful and ubiquitous tools, AI models exhibit one detrimential
characteristic: a performance-transparency trade-off. This describes the fact
that the more complex a model's inner workings, the less clear it is how its
predictions or decisions were achieved. But, especially considering Machine
Learning (ML) methods like Reinforcement Learning (RL) where the system learns
autonomously, the necessity to understand the underlying reasoning for their
decisions becomes apparent. Since, to the best of our knowledge, there exists
no single work offering an overview of Explainable Reinforcement Learning (XRL)
methods, this survey attempts to address this gap. We give a short summary of
the problem, a definition of important terms, and offer a classification and
assessment of current XRL methods. We found that a) the majority of XRL methods
function by mimicking and simplifying a complex model instead of designing an
inherently simple one, and b) XRL (and XAI) methods often neglect to consider
the human side of the equation, not taking into account research from related
fields like psychology or philosophy. Thus, an interdisciplinary effort is
needed to adapt the generated explanations to a (non-expert) human user in
order to effectively progress in the field of XRL and XAI in general. | [
"cs.LG",
"stat.ML",
"A.1"
] |
3D point clouds play pivotal roles in various safety-critical applications,
such as autonomous driving, which desires the underlying deep neural networks
to be robust to adversarial perturbations. Though a few defenses against
adversarial point cloud classification have been proposed, it remains unknown
whether they are truly robust to adaptive attacks. To this end, we perform the
first security analysis of state-of-the-art defenses and design adaptive
evaluations on them. Our 100% adaptive attack success rates show that current
countermeasures are still vulnerable. Since adversarial training (AT) is
believed as the most robust defense, we present the first in-depth study
showing how AT behaves in point cloud classification and identify that the
required symmetric function (pooling operation) is paramount to the 3D model's
robustness under AT. Through our systematic analysis, we find that the
default-used fixed pooling (e.g., MAX pooling) generally weakens AT's
effectiveness in point cloud classification. Interestingly, we further discover
that sorting-based parametric pooling can significantly improve the models'
robustness. Based on above insights, we propose DeepSym, a deep symmetric
pooling operation, to architecturally advance the robustness to 47.0% under AT
without sacrificing nominal accuracy, outperforming the original design and a
strong baseline by 28.5% ($\sim 2.6 \times$) and 6.5%, respectively, in
PointNet. | [
"cs.LG",
"cs.AI",
"cs.CR"
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.