text
stringlengths 29
3.31k
| label
sequencelengths 1
11
|
|---|---|
Forecasting a time series from multivariate predictors constitutes a
challenging problem, especially using model-free approaches. Most techniques,
such as nearest-neighbor prediction, quickly suffer from the curse of
dimensionality and overfitting for more than a few predictors which has limited
their application mostly to the univariate case. Therefore, selection
strategies are needed that harness the available information as efficiently as
possible. Since often the right combination of predictors matters, ideally all
subsets of possible predictors should be tested for their predictive power, but
the exponentially growing number of combinations makes such an approach
computationally prohibitive. Here a prediction scheme that overcomes this
strong limitation is introduced utilizing a causal pre-selection step which
drastically reduces the number of possible predictors to the most predictive
set of causal drivers making a globally optimal search scheme tractable. The
information-theoretic optimality is derived and practical selection criteria
are discussed. As demonstrated for multivariate nonlinear stochastic delay
processes, the optimal scheme can even be less computationally expensive than
commonly used sub-optimal schemes like forward selection. The method suggests a
general framework to apply the optimal model-free approach to select variables
and subsequently fit a model to further improve a prediction or learn
statistical dependencies. The performance of this framework is illustrated on a
climatological index of El Ni\~no Southern Oscillation. | [
"stat.ML",
"stat.ME"
] |
In this paper, we present a new approach to interpret deep learning models.
By coupling mutual information with network science, we explore how information
flows through feedforward networks. We show that efficiently approximating
mutual information allows us to create an information measure that quantifies
how much information flows between any two neurons of a deep learning model. To
that end, we propose NIF, Neural Information Flow, a technique for codifying
information flow that exposes deep learning model internals and provides
feature attributions. | [
"cs.LG",
"stat.ML"
] |
The goal of imitation learning (IL) is to learn a good policy from
high-quality demonstrations. However, the quality of demonstrations in reality
can be diverse, since it is easier and cheaper to collect demonstrations from a
mix of experts and amateurs. IL in such situations can be challenging,
especially when the level of demonstrators' expertise is unknown. We propose a
new IL method called \underline{v}ariational \underline{i}mitation
\underline{l}earning with \underline{d}iverse-quality demonstrations (VILD),
where we explicitly model the level of demonstrators' expertise with a
probabilistic graphical model and estimate it along with a reward function. We
show that a naive approach to estimation is not suitable to large state and
action spaces, and fix its issues by using a variational approach which can be
easily implemented using existing reinforcement learning methods. Experiments
on continuous-control benchmarks demonstrate that VILD outperforms
state-of-the-art methods. Our work enables scalable and data-efficient IL under
more realistic settings than before. | [
"cs.LG",
"stat.ML"
] |
Light field photography has been studied thoroughly in recent years. One of
its drawbacks is the need for multi-lens in the imaging. To compensate that,
compressed light field photography has been proposed to tackle the trade-offs
between the spatial and angular resolutions. It obtains by only one lens, a
compressed version of the regular multi-lens system. The acquisition system
consists of a dedicated hardware followed by a decompression algorithm, which
usually suffers from high computational time. In this work, we propose a
computationally efficient neural network that recovers a high-quality color
light field from a single coded image. Unlike previous works, we compress the
color channels as well, removing the need for a CFA in the imaging system. Our
approach outperforms existing solutions in terms of recovery quality and
computational complexity. We propose also a neural network for depth map
extraction based on the decompressed light field, which is trained in an
unsupervised manner without the ground truth depth map. | [
"cs.CV"
] |
We address the problem of semantic nighttime image segmentation and improve
the state-of-the-art, by adapting daytime models to nighttime without using
nighttime annotations. Moreover, we design a new evaluation framework to
address the substantial uncertainty of semantics in nighttime images. Our
central contributions are: 1) a curriculum framework to gradually adapt
semantic segmentation models from day to night through progressively darker
times of day, exploiting cross-time-of-day correspondences between daytime
images from a reference map and dark images to guide the label inference in the
dark domains; 2) a novel uncertainty-aware annotation and evaluation framework
and metric for semantic segmentation, including image regions beyond human
recognition capability in the evaluation in a principled fashion; 3) the Dark
Zurich dataset, comprising 2416 unlabeled nighttime and 2920 unlabeled twilight
images with correspondences to their daytime counterparts plus a set of 201
nighttime images with fine pixel-level annotations created with our protocol,
which serves as a first benchmark for our novel evaluation. Experiments show
that our map-guided curriculum adaptation significantly outperforms
state-of-the-art methods on nighttime sets both for standard metrics and our
uncertainty-aware metric. Furthermore, our uncertainty-aware evaluation reveals
that selective invalidation of predictions can improve results on data with
ambiguous content such as our benchmark and profit safety-oriented applications
involving invalid inputs. | [
"cs.CV"
] |
In the field of Natural Language Processing (NLP), we revisit the well-known
word embedding algorithm word2vec. Word embeddings identify words by vectors
such that the words' distributional similarity is captured. Unexpectedly,
besides semantic similarity even relational similarity has been shown to be
captured in word embeddings generated by word2vec, whence two questions arise.
Firstly, which kind of relations are representable in continuous space and
secondly, how are relations built. In order to tackle these questions we
propose a bottom-up point of view. We call generating input text for which
word2vec outputs target relations solving the Corpus Replication Task. Deeming
generalizations of this approach to any set of relations possible, we expect
solving of the Corpus Replication Task to provide partial answers to the
questions. | [
"cs.LG",
"cs.CL",
"stat.ML"
] |
Motivated by high-stakes decision-making domains like personalized medicine
where user information is inherently sensitive, we design privacy preserving
exploration policies for episodic reinforcement learning (RL). We first provide
a meaningful privacy formulation using the notion of joint differential privacy
(JDP)--a strong variant of differential privacy for settings where each user
receives their own sets of output (e.g., policy recommendations). We then
develop a private optimism-based learning algorithm that simultaneously
achieves strong PAC and regret bounds, and enjoys a JDP guarantee. Our
algorithm only pays for a moderate privacy cost on exploration: in comparison
to the non-private bounds, the privacy parameter only appears in lower-order
terms. Finally, we present lower bounds on sample complexity and regret for
reinforcement learning subject to JDP. | [
"cs.LG",
"cs.CR",
"stat.ML"
] |
Previous work in hierarchical reinforcement learning has faced a dilemma:
either ignore the values of different possible exit states from a subroutine,
thereby risking suboptimal behavior, or represent those values explicitly
thereby incurring a possibly large representation cost because exit values
refer to nonlocal aspects of the world (i.e., all subsequent rewards). This
paper shows that, in many cases, one can avoid both of these problems. The
solution is based on recursively decomposing the exit value function in terms
of Q-functions at higher levels of the hierarchy. This leads to an intuitively
appealing runtime architecture in which a parent subroutine passes to its child
a value function on the exit states and the child reasons about how its choices
affect the exit value. We also identify structural conditions on the value
function and transition distributions that allow much more concise
representations of exit state distributions, leading to further state
abstraction. In essence, the only variables whose exit values need be
considered are those that the parent cares about and the child affects. We
demonstrate the utility of our algorithms on a series of increasingly complex
environments. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Lidar based 3D object detection is inevitable for autonomous driving, because
it directly links to environmental understanding and therefore builds the base
for prediction and motion planning. The capacity of inferencing highly sparse
3D data in real-time is an ill-posed problem for lots of other application
areas besides automated vehicles, e.g. augmented reality, personal robotics or
industrial automation. We introduce Complex-YOLO, a state of the art real-time
3D object detection network on point clouds only. In this work, we describe a
network that expands YOLOv2, a fast 2D standard object detector for RGB images,
by a specific complex regression strategy to estimate multi-class 3D boxes in
Cartesian space. Thus, we propose a specific Euler-Region-Proposal Network
(E-RPN) to estimate the pose of the object by adding an imaginary and a real
fraction to the regression network. This ends up in a closed complex space and
avoids singularities, which occur by single angle estimations. The E-RPN
supports to generalize well during training. Our experiments on the KITTI
benchmark suite show that we outperform current leading methods for 3D object
detection specifically in terms of efficiency. We achieve state of the art
results for cars, pedestrians and cyclists by being more than five times faster
than the fastest competitor. Further, our model is capable of estimating all
eight KITTI-classes, including Vans, Trucks or sitting pedestrians
simultaneously with high accuracy. | [
"cs.CV"
] |
Recently, there is an increasing demand for automatically detecting
anatomical landmarks which provide rich structural information to facilitate
subsequent medical image analysis. Current methods related to this task often
leverage the power of deep neural networks, while a major challenge in fine
tuning such models in medical applications arises from insufficient number of
labeled samples. To address this, we propose to regularize the knowledge
transfer across source and target tasks through cross-task representation
learning. The proposed method is demonstrated for extracting facial anatomical
landmarks which facilitate the diagnosis of fetal alcohol syndrome. The source
and target tasks in this work are face recognition and landmark detection,
respectively. The main idea of the proposed method is to retain the feature
representations of the source model on the target task data, and to leverage
them as an additional source of supervisory signals for regularizing the target
model learning, thereby improving its performance under limited training
samples. Concretely, we present two approaches for the proposed representation
learning by constraining either final or intermediate model features on the
target model. Experimental results on a clinical face image dataset demonstrate
that the proposed approach works well with few labeled data, and outperforms
other compared approaches. | [
"cs.CV",
"cs.LG",
"eess.IV"
] |
A regression-based BNN model is proposed to predict spatiotemporal quantities
like hourly rider demand with calibrated uncertainties. The main contributions
of this paper are (i) A feed-forward deterministic neural network (DetNN)
architecture that predicts cyclical time series data with sensitivity to
anomalous forecasting events; (ii) A Bayesian framework applying SVGD to train
large neural networks for such tasks, capable of producing time series
predictions as well as measures of uncertainty surrounding the predictions.
Experiments show that the proposed BNN reduces average estimation error by 10%
across 8 U.S. cities compared to a fine-tuned multilayer perceptron (MLP), and
4% better than the same network architecture trained without SVGD. | [
"cs.LG",
"stat.ML"
] |
Monitoring the dynamics of traffic in major corridors can provide invaluable
insight for traffic planning purposes. An important requirement for this
monitoring is the availability of methods to automatically detect major traffic
events and to annotate the abundance of travel data. This paper introduces a
machine learning based approach for reliable detection and characterization of
highway traffic congestion events from hundreds of hours of traffic speed data.
Indeed, the proposed approach is a generic approach for detection of changes in
any given time series, which is the wireless traffic sensor data in the present
study. The speed data is initially time-windowed by a ten-hour long sliding
window and fed into three Neural Networks that are used to detect the existence
and duration of congestion events (slowdowns) in each window. The sliding
window captures each slowdown event multiple times and results in increased
confidence in congestion detection. The training and parameter tuning are
performed on 17,483 hours of data that includes 168 slowdown events. This data
is collected and labeled as part of the ongoing probe data validation studies
at the Center for Advanced Transportation Technologies (CATT) at the University
of Maryland. The Neural networks are carefully trained to reduce the chances of
over-fitting to the training data. The experimental results show that this
approach is able to successfully detect most of the congestion events, while
significantly outperforming a heuristic rule-based approach. Moreover, the
proposed approach is shown to be more accurate in estimation of the start-time
and end-time of the congestion events. | [
"cs.CV",
"cs.LG",
"eess.SP"
] |
Digital Twins have been described as beneficial in many areas, such as
virtual commissioning, fault prediction or reconfiguration planning. Equipping
Digital Twins with artificial intelligence functionalities can greatly expand
those beneficial applications or open up altogether new areas of application,
among them cross-phase industrial transfer learning. In the context of machine
learning, transfer learning represents a set of approaches that enhance
learning new tasks based upon previously acquired knowledge. Here, knowledge is
transferred from one lifecycle phase to another in order to reduce the amount
of data or time needed to train a machine learning algorithm. Looking at common
challenges in developing and deploying industrial machinery with deep learning
functionalities, embracing this concept would offer several advantages: Using
an intelligent Digital Twin, learning algorithms can be designed, configured
and tested in the design phase before the physical system exists and real data
can be collected. Once real data becomes available, the algorithms must merely
be fine-tuned, significantly speeding up commissioning and reducing the
probability of costly modifications. Furthermore, using the Digital Twin's
simulation capabilities virtually injecting rare faults in order to train an
algorithm's response or using reinforcement learning, e.g. to teach a robot,
become practically feasible. This article presents several cross-phase
industrial transfer learning use cases utilizing intelligent Digital Twins. A
real cyber physical production system consisting of an automated welding
machine and an automated guided vehicle equipped with a robot arm is used to
illustrate the respective benefits. | [
"cs.LG"
] |
We introduce DatasetGAN: an automatic procedure to generate massive datasets
of high-quality semantically segmented images requiring minimal human effort.
Current deep networks are extremely data-hungry, benefiting from training on
large-scale datasets, which are time consuming to annotate. Our method relies
on the power of recent GANs to generate realistic images. We show how the GAN
latent code can be decoded to produce a semantic segmentation of the image.
Training the decoder only needs a few labeled examples to generalize to the
rest of the latent space, resulting in an infinite annotated dataset generator!
These generated datasets can then be used for training any computer vision
architecture just as real datasets are. As only a few images need to be
manually segmented, it becomes possible to annotate images in extreme detail
and generate datasets with rich object and part segmentations. To showcase the
power of our approach, we generated datasets for 7 image segmentation tasks
which include pixel-level labels for 34 human face parts, and 32 car parts. Our
approach outperforms all semi-supervised baselines significantly and is on par
with fully supervised methods, which in some cases require as much as 100x more
annotated data as our method. | [
"cs.CV"
] |
In this paper, we propose a novel technique for generating images in the 3D
domain from images with high degree of geometrical transformations. By
coalescing two popular concurrent methods that have seen rapid ascension to the
machine learning zeitgeist in recent years: GANs (Goodfellow et. al.) and
Capsule networks (Sabour, Hinton et. al.) - we present: \textbf{CapsGAN}. We
show that CapsGAN performs better than or equal to traditional CNN based GANs
in generating images with high geometric transformations using rotated MNIST.
In the process, we also show the efficacy of using capsules architecture in the
GANs domain. Furthermore, we tackle the Gordian Knot in training GANs - the
performance control and training stability by experimenting with using
Wasserstein distance (gradient clipping, penalty) and Spectral Normalization.
The experimental findings of this paper should propel the application of
capsules and GANs in the still exciting and nascent domain of 3D image
generation, and plausibly video (frame) generation. | [
"cs.CV",
"cs.LG"
] |
In recent years, there has been a growing interest in applying convolutional
neural networks (CNNs) to low-level vision tasks such as denoising and
super-resolution. Due to the coherent nature of the image formation process,
optical coherence tomography (OCT) images are inevitably affected by noise.
This paper proposes a new method named the multi-input fully-convolutional
networks (MIFCN) for denoising of OCT images. In contrast to recently proposed
natural image denoising CNNs, the proposed architecture allows the exploitation
of high degrees of correlation and complementary information among neighboring
OCT images through pixel by pixel fusion of multiple FCNs. The parameters of
the proposed multi-input architecture are learned by considering the
consistency between the overall output and the contribution of each input
image. The proposed MIFCN method is compared with the state-of-the-art
denoising methods adopted on OCT images of normal and age-related macular
degeneration eyes in a quantitative and qualitative manner. | [
"cs.CV"
] |
Safe and proactive planning in robotic systems generally requires accurate
predictions of the environment. Prior work on environment prediction applied
video frame prediction techniques to bird's-eye view environment
representations, such as occupancy grids. ConvLSTM-based frameworks used
previously often result in significant blurring and vanishing of moving
objects, thus hindering their applicability for use in safety-critical
applications. In this work, we propose two extensions to the ConvLSTM to
address these issues. We present the Temporal Attention Augmented ConvLSTM
(TAAConvLSTM) and Self-Attention Augmented ConvLSTM (SAAConvLSTM) frameworks
for spatiotemporal occupancy prediction, and demonstrate improved performance
over baseline architectures on the real-world KITTI and Waymo datasets. | [
"cs.CV",
"cs.AI",
"cs.LG",
"cs.RO",
"I.2.9; I.2.10"
] |
Traditional methods for demand forecasting only focus on modeling the
temporal dependency. However, forecasting on spatio-temporal data requires
modeling of complex nonlinear relational and spatial dependencies. In addition,
dynamic contextual information can have a significant impact on the demand
values, and therefore needs to be captured. For example, in a bike-sharing
system, bike usage can be impacted by weather. Existing methods assume the
contextual impact is fixed. However, we note that the contextual impact evolves
over time. We propose a novel context integrated relational model, Context
Integrated Graph Neural Network (CIGNN), which leverages the temporal,
relational, spatial, and dynamic contextual dependencies for multi-step ahead
demand forecasting. Our approach considers the demand network over various
geographical locations and represents the network as a graph. We define a
demand graph, where nodes represent demand time-series, and context graphs (one
for each type of context), where nodes represent contextual time-series.
Assuming that various contexts evolve and have a dynamic impact on the
fluctuation of demand, our proposed CIGNN model employs a fusion mechanism that
jointly learns from all available types of contextual information. To the best
of our knowledge, this is the first approach that integrates dynamic contexts
with graph neural networks for spatio-temporal demand forecasting, thereby
increasing prediction accuracy. We present empirical results on two real-world
datasets, demonstrating that CIGNN consistently outperforms state-of-the-art
baselines, in both periodic and irregular time-series networks. | [
"cs.LG",
"stat.ML"
] |
Low-cost particulate matter sensors are transforming air quality monitoring
because they have lower costs and greater mobility as compared to reference
monitors. Calibration of these low-cost sensors requires training data from
co-deployed reference monitors. Machine Learning based calibration gives better
performance than conventional techniques, but requires a large amount of
training data from the sensor, to be calibrated, co-deployed with a reference
monitor. In this work, we propose novel transfer learning methods for quick
calibration of sensors with minimal co-deployment with reference monitors.
Transfer learning utilizes a large amount of data from other sensors along with
a limited amount of data from the target sensor. Our extensive experimentation
finds the proposed Model-Agnostic- Meta-Learning (MAML) based transfer learning
method to be the most effective over other competitive baselines. | [
"cs.LG",
"eess.SP"
] |
Detecting communities on graphs has received significant interest in recent
literature. Current state-of-the-art community embedding approach called
\textit{ComE} tackles this problem by coupling graph embedding with community
detection. Considering the success of hyperbolic representations of
graph-structured data in last years, an ongoing challenge is to set up a
hyperbolic approach for the community detection problem. The present paper
meets this challenge by introducing a Riemannian equivalent of \textit{ComE}.
Our proposed approach combines hyperbolic embeddings with Riemannian K-means or
Riemannian mixture models to perform community detection. We illustrate the
usefulness of this framework through several experiments on real-world social
networks and comparisons with \textit{ComE} and recent hyperbolic-based
classification approaches. | [
"cs.LG",
"stat.ML"
] |
Transformers have shown impressive performance in various natural language
processing and computer vision tasks, due to the capability of modeling
long-range dependencies. Recent progress has demonstrated to combine such
transformers with CNN-based semantic image segmentation models is very
promising. However, it is not well studied yet on how well a pure transformer
based approach can achieve for image segmentation. In this work, we explore a
novel framework for semantic image segmentation, which is encoder-decoder based
Fully Transformer Networks (FTN). Specifically, we first propose a Pyramid
Group Transformer (PGT) as the encoder for progressively learning hierarchical
features, while reducing the computation complexity of the standard visual
transformer(ViT). Then, we propose a Feature Pyramid Transformer (FPT) to fuse
semantic-level and spatial-level information from multiple levels of the PGT
encoder for semantic image segmentation. Surprisingly, this simple baseline can
achieve new state-of-the-art results on multiple challenging semantic
segmentation benchmarks, including PASCAL Context, ADE20K and COCO-Stuff. The
source code will be released upon the publication of this work. | [
"cs.CV"
] |
Shadow detection and shadow removal are fundamental and challenging tasks,
requiring an understanding of the global image semantics. This paper presents a
novel deep neural network design for shadow detection and removal by analyzing
the spatial image context in a direction-aware manner. To achieve this, we
first formulate the direction-aware attention mechanism in a spatial recurrent
neural network (RNN) by introducing attention weights when aggregating spatial
context features in the RNN. By learning these weights through training, we can
recover direction-aware spatial context (DSC) for detecting and removing
shadows. This design is developed into the DSC module and embedded in a
convolutional neural network (CNN) to learn the DSC features at different
levels. Moreover, we design a weighted cross entropy loss to make effective the
training for shadow detection and further adopt the network for shadow removal
by using a Euclidean loss function and formulating a color transfer function to
address the color and luminosity inconsistencies in the training pairs. We
employed two shadow detection benchmark datasets and two shadow removal
benchmark datasets, and performed various experiments to evaluate our method.
Experimental results show that our method performs favorably against the
state-of-the-art methods for both shadow detection and shadow removal. | [
"cs.CV"
] |
Generative Adversarial Networks (GANs) are a class of generative algorithms
that have been shown to produce state-of-the art samples, especially in the
domain of image creation. The fundamental principle of GANs is to approximate
the unknown distribution of a given data set by optimizing an objective
function through an adversarial game between a family of generators and a
family of discriminators. In this paper, we offer a better theoretical
understanding of GANs by analyzing some of their mathematical and statistical
properties. We study the deep connection between the adversarial principle
underlying GANs and the Jensen-Shannon divergence, together with some
optimality characteristics of the problem. An analysis of the role of the
discriminator family via approximation arguments is also provided. In addition,
taking a statistical point of view, we study the large sample properties of the
estimated distribution and prove in particular a central limit theorem. Some of
our results are illustrated with simulated examples. | [
"stat.ML",
"cs.LG"
] |
Object skeleton is a useful cue for object detection, complementary to the
object contour, as it provides a structural representation to describe the
relationship among object parts. While object skeleton extraction in natural
images is a very challenging problem, as it requires the extractor to be able
to capture both local and global image context to determine the intrinsic scale
of each skeleton pixel. Existing methods rely on per-pixel based multi-scale
feature computation, which results in difficult modeling and high time
consumption. In this paper, we present a fully convolutional network with
multiple scale-associated side outputs to address this problem. By observing
the relationship between the receptive field sizes of the sequential stages in
the network and the skeleton scales they can capture, we introduce a
scale-associated side output to each stage. We impose supervision to different
stages by guiding the scale-associated side outputs toward groundtruth
skeletons of different scales. The responses of the multiple scale-associated
side outputs are then fused in a scale-specific way to localize skeleton pixels
with multiple scales effectively. Our method achieves promising results on two
skeleton extraction datasets, and significantly outperforms other competitors. | [
"cs.CV"
] |
Representation learning has overcome the often arduous and manual
featurization of networks through (unsupervised) feature learning as it results
in embeddings that can apply to a variety of downstream learning tasks. The
focus of representation learning on graphs has focused mainly on shallow
(node-centric) or deep (graph-based) learning approaches. While there have been
approaches that work on homogeneous and heterogeneous networks with multi-typed
nodes and edges, there is a gap in learning edge representations. This paper
proposes a novel unsupervised inductive method called AttrE2Vec, which learns a
low-dimensional vector representation for edges in attributed networks. It
systematically captures the topological proximity, attributes affinity, and
feature similarity of edges. Contrary to current advances in edge embedding
research, our proposal extends the body of methods providing representations
for edges, capturing graph attributes in an inductive and unsupervised manner.
Experimental results show that, compared to contemporary approaches, our method
builds more powerful edge vector representations, reflected by higher quality
measures (AUC, accuracy) in downstream tasks as edge classification and edge
clustering. It is also confirmed by analyzing low-dimensional embedding
projections. | [
"cs.LG"
] |
Reinforcement learning methods trained on few environments rarely learn
policies that generalize to unseen environments. To improve generalization, we
incorporate the inherent sequential structure in reinforcement learning into
the representation learning process. This approach is orthogonal to recent
approaches, which rarely exploit this structure explicitly. Specifically, we
introduce a theoretically motivated policy similarity metric (PSM) for
measuring behavioral similarity between states. PSM assigns high similarity to
states for which the optimal policies in those states as well as in future
states are similar. We also present a contrastive representation learning
procedure to embed any state similarity metric, which we instantiate with PSM
to obtain policy similarity embeddings (PSEs). We demonstrate that PSEs improve
generalization on diverse benchmarks, including LQR with spurious correlations,
a jumping task from pixels, and Distracting DM Control Suite. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Depth estimation from a stereo image pair has become one of the most explored
applications in computer vision, with most of the previous methods relying on
fully supervised learning settings. However, due to the difficulty in acquiring
accurate and scalable ground truth data, the training of fully supervised
methods is challenging. As an alternative, self-supervised methods are becoming
more popular to mitigate this challenge. In this paper, we introduce the H-Net,
a deep-learning framework for unsupervised stereo depth estimation that
leverages epipolar geometry to refine stereo matching. For the first time, a
Siamese autoencoder architecture is used for depth estimation which allows
mutual information between the rectified stereo images to be extracted. To
enforce the epipolar constraint, the mutual epipolar attention mechanism has
been designed which gives more emphasis to correspondences of features which
lie on the same epipolar line while learning mutual information between the
input stereo pair. Stereo correspondences are further enhanced by incorporating
semantic information to the proposed attention mechanism. More specifically,
the optimal transport algorithm is used to suppress attention and eliminate
outliers in areas not visible in both cameras. Extensive experiments on
KITTI2015 and Cityscapes show that our method outperforms the state-ofthe-art
unsupervised stereo depth estimation methods while closing the gap with the
fully supervised approaches. | [
"cs.CV"
] |
The placenta is a complex organ, playing multiple roles during fetal
development. Very little is known about the association between placental
morphological abnormalities and fetal physiology. In this work, we present an
open sourced, computationally tractable deep learning pipeline to analyse
placenta histology at the level of the cell. By utilising two deep
Convolutional Neural Network architectures and transfer learning, we can
robustly localise and classify placental cells within five classes with an
accuracy of 89%. Furthermore, we learn deep embeddings encoding phenotypic
knowledge that is capable of both stratifying five distinct cell populations
and learn intraclass phenotypic variance. We envisage that the automation of
this pipeline to population scale studies of placenta histology has the
potential to improve our understanding of basic cellular placental biology and
its variations, particularly its role in predicting adverse birth outcomes. | [
"cs.CV"
] |
We propose a stereo vision-based approach for tracking the camera ego-motion
and 3D semantic objects in dynamic autonomous driving scenarios. Instead of
directly regressing the 3D bounding box using end-to-end approaches, we propose
to use the easy-to-labeled 2D detection and discrete viewpoint classification
together with a light-weight semantic inference method to obtain rough 3D
object measurements. Based on the object-aware-aided camera pose tracking which
is robust in dynamic environments, in combination with our novel dynamic object
bundle adjustment (BA) approach to fuse temporal sparse feature correspondences
and the semantic 3D measurement model, we obtain 3D object pose, velocity and
anchored dynamic point cloud estimation with instance accuracy and temporal
consistency. The performance of our proposed method is demonstrated in diverse
scenarios. Both the ego-motion estimation and object localization are compared
with the state-of-of-the-art solutions. | [
"cs.CV"
] |
Endogeneity bias and instrument variable validation have always been
important topics in statistics and econometrics. In the era of big data, such
issues typically combine with dimensionality issues and, hence, require even
more attention. In this paper, we merge two well-known tools from machine
learning and biostatistics---variable selection algorithms and probablistic
graphs---to estimate house prices and the corresponding causal structure using
2010 data on Sydney. The estimation uses a 200-gigabyte ultrahigh dimensional
database consisting of local school data, GIS information, census data, house
characteristics and other socio-economic records. Using "big data", we show
that it is possible to perform a data-driven instrument selection efficiently
and purge out the invalid instruments. Our approach improves the sparsity of
variable selection, stability and robustness in the presence of high
dimensionality, complicated causal structures and the consequent
multicollinearity, and recovers a sparse and intuitive causal structure. The
approach also reveals an efficiency and effectiveness in endogeneity detection,
instrument validation, weak instrument pruning and the selection of valid
instruments. From the perspective of machine learning, the estimation results
both align with and confirms the facts of Sydney house market, the classical
economic theories and the previous findings of simultaneous equations modeling.
Moreover, the estimation results are consistent with and supported by classical
econometric tools such as two-stage least square regression and different
instrument tests. All the code may be found at
\url{https://github.com/isaac2math/solar_graph_learning}. | [
"stat.ML",
"cs.LG",
"stat.AP"
] |
In person re-identification, extracting part-level features from person
images has been verified to be crucial. Most of existing CNN-based methods only
locate the human parts coarsely, or rely on pre-trained human parsing models
and fail in locating the identifiable non-human parts (e.g., knapsack). In this
paper, we introduce an alignment scheme in Transformer architecture for the
first time and propose the Auto-Aligned Transformer (AAformer) to automatically
locate both the human parts and non-human ones at patch-level. We introduce the
"part tokens", which are learnable vectors, to extract part features in
Transformer. A part token only interacts with a local subset of patches in
self-attention and learns to be the part representation. To adaptively group
the image patches into different subsets, we design the Auto-Alignment.
Auto-Alignment employs a fast variant of Optimal Transport algorithm to online
cluster the patch embeddings into several groups with the part tokens as their
prototypes. We harmoniously integrate the part alignment into the
self-attention and the output part tokens can be directly used for retrieval.
Extensive experiments validate the effectiveness of part tokens and the
superiority of AAformer over various state-of-the-art methods. | [
"cs.CV"
] |
With the decreasing cost of data collection, the space of variables or
features that can be used to characterize a particular predictor of interest
continues to grow exponentially. Therefore, identifying the most characterizing
features that minimizes the variance without jeopardizing the bias of our
models is critical to successfully training a machine learning model. In
addition, identifying such features is critical for interpretability,
prediction accuracy and optimal computation cost. While statistical methods
such as subset selection, shrinkage, dimensionality reduction have been applied
in selecting the best set of features, some other approaches in literature have
approached feature selection task as a search problem where each state in the
search space is a possible feature subset. In this paper, we solved the feature
selection problem using Reinforcement Learning. Formulating the state space as
a Markov Decision Process (MDP), we used Temporal Difference (TD) algorithm to
select the best subset of features. Each state was evaluated using a robust and
low cost classifier algorithm which could handle any non-linearities in the
dataset. | [
"cs.LG",
"stat.ML"
] |
Cytology is a low-cost and non-invasive diagnostic procedure employed to
support the diagnosis of a broad range of pathologies. Computer Vision
technologies, by automatically generating quantitative and objective
descriptions of examinations' contents, can help minimize the chances of
misdiagnoses and shorten the time required for analysis. To identify the
state-of-art of computer vision techniques currently applied to cytology, we
conducted a Systematic Literature Review. We analyzed papers published in the
last 5 years. The initial search was executed in September 2020 and resulted in
431 articles. After applying the inclusion/exclusion criteria, 157 papers
remained, which we analyzed to build a picture of the tendencies and problems
present in this research area, highlighting the computer vision methods,
staining techniques, evaluation metrics, and the availability of the used
datasets and computer code. As a result, we identified that the most used
methods in the analyzed works are deep learning-based (70 papers), while fewer
works employ classic computer vision only (101 papers). The most recurrent
metric used for classification and object detection was the accuracy (33 papers
and 5 papers), while for segmentation it was the Dice Similarity Coefficient
(38 papers). Regarding staining techniques, Papanicolaou was the most employed
one (130 papers), followed by H&E (20 papers) and Feulgen (5 papers). Twelve of
the datasets used in the papers are publicly available, with the DTU/Herlev
dataset being the most used one. We conclude that there still is a lack of
high-quality datasets for many types of stains and most of the works are not
mature enough to be applied in a daily clinical diagnostic routine. We also
identified a growing tendency towards adopting deep learning-based approaches
as the methods of choice. | [
"cs.CV",
"eess.IV"
] |
Graph Neural Networks (GNNs) are deep learning methods which provide the
current state of the art performance in node classification tasks. GNNs often
assume homophily -- neighboring nodes having similar features and labels--, and
therefore may not be at their full potential when dealing with non-homophilic
graphs. In this work, we focus on addressing this limitation and enable Graph
Attention Networks (GAT), a commonly used variant of GNNs, to explore the
structural information within each graph locality. Inspired by the positional
encoding in the Transformers, we propose a framework, termed Graph Attentional
Networks with Positional Embeddings (GAT-POS), to enhance GATs with positional
embeddings which capture structural and positional information of the nodes in
the graph. In this framework, the positional embeddings are learned by a model
predictive of the graph context, plugged into an enhanced GAT architecture,
which is able to leverage both the positional and content information of each
node. The model is trained jointly to optimize for the task of node
classification as well as the task of predicting graph context. Experimental
results show that GAT-POS reaches remarkable improvement compared to strong GNN
baselines and recent structural embedding enhanced GNNs on non-homophilic
graphs. | [
"cs.LG"
] |
Although numerous improvements have been made in the field of image
segmentation using convolutional neural networks, the majority of these
improvements rely on training with larger datasets, model architecture
modifications, novel loss functions, and better optimizers. In this paper, we
propose a new segmentation performance boosting paradigm that relies on
optimally modifying the network's input instead of the network itself. In
particular, we leverage the gradients of a trained segmentation network with
respect to the input to transfer it to a space where the segmentation accuracy
improves. We test the proposed method on three publicly available medical image
segmentation datasets: the ISIC 2017 Skin Lesion Segmentation dataset, the
Shenzhen Chest X-Ray dataset, and the CVC-ColonDB dataset, for which our method
achieves improvements of 5.8%, 0.5%, and 4.8% in the average Dice scores,
respectively. | [
"cs.CV"
] |
This paper introduces MDP homomorphic networks for deep reinforcement
learning. MDP homomorphic networks are neural networks that are equivariant
under symmetries in the joint state-action space of an MDP. Current approaches
to deep reinforcement learning do not usually exploit knowledge about such
structure. By building this prior knowledge into policy and value networks
using an equivariance constraint, we can reduce the size of the solution space.
We specifically focus on group-structured symmetries (invertible
transformations). Additionally, we introduce an easy method for constructing
equivariant network layers numerically, so the system designer need not solve
the constraints by hand, as is typically done. We construct MDP homomorphic
MLPs and CNNs that are equivariant under either a group of reflections or
rotations. We show that such networks converge faster than unstructured
baselines on CartPole, a grid world and Pong. | [
"cs.LG",
"stat.ML"
] |
The tracking-by-detection framework usually consist of two stages: drawing
samples around the target object in the first stage and classifying each sample
as the target object or background in the second stage. Current popular
trackers based on tracking-by-detection framework typically draw samples in the
raw image as the inputs of deep convolution networks in the first stage, which
usually results in high computational burden and low running speed. In this
paper, we propose a new visual tracking method using sampling deep
convolutional features to address this problem. Only one cropped image around
the target object is input into the designed deep convolution network and the
samples is sampled on the feature maps of the network by spatial bilinear
resampling. In addition, a generative adversarial network is integrated into
our network framework to augment positive samples and improve the tracking
performance. Extensive experiments on benchmark datasets demonstrate that the
proposed method achieves a comparable performance to state-of-the-art trackers
and accelerates tracking-by-detection trackers based on raw-image samples
effectively. | [
"cs.CV"
] |
Learned pointcloud representations do not generalize well with an increase in
distance to the sensor. For example, at a range greater than 60 meters, the
sparsity of lidar pointclouds reaches to a point where even humans cannot
discern object shapes from each other. However, this distance should not be
considered very far for fast-moving vehicles: A vehicle can traverse 60 meters
under two seconds while moving at 70 mph. For safe and robust driving
automation, acute 3D object detection at these ranges is indispensable. Against
this backdrop, we introduce faraway-frustum: a novel fusion strategy for
detecting faraway objects. The main strategy is to depend solely on the 2D
vision for recognizing object class, as object shape does not change
drastically with an increase in depth, and use pointcloud data for object
localization in the 3D space for faraway objects. For closer objects, we use
learned pointcloud representations instead, following state-of-the-art. This
strategy alleviates the main shortcoming of object detection with learned
pointcloud representations. Experiments on the KITTI dataset demonstrate that
our method outperforms state-of-the-art by a considerable margin for faraway
object detection in bird's-eye-view and 3D. Our code is open-source and
publicly available: https://github.com/dongfang-steven-yang/faraway-frustum. | [
"cs.CV"
] |
This paper presents an approach to address data scarcity problems in
underwater image datasets for visual detection of marine debris. The proposed
approach relies on a two-stage variational autoencoder (VAE) and a binary
classifier to evaluate the generated imagery for quality and realism. From the
images generated by the two-stage VAE, the binary classifier selects "good
quality" images and augments the given dataset with them. Lastly, a multi-class
classifier is used to evaluate the impact of the augmentation process by
measuring the accuracy of an object detector trained on combinations of real
and generated trash images. Our results show that the classifier trained with
the augmented data outperforms the one trained only with the real data. This
approach will not only be valid for the underwater trash classification problem
presented in this paper, but it will also be useful for any data-dependent task
for which collecting more images is challenging or infeasible. | [
"cs.CV",
"cs.RO"
] |
Motivated by the pursuit of a systematic computational and algorithmic
understanding of Generative Adversarial Networks (GANs), we present a simple
yet unified non-asymptotic local convergence theory for smooth two-player
games, which subsumes several discrete-time gradient-based saddle point
dynamics. The analysis reveals the surprising nature of the off-diagonal
interaction term as both a blessing and a curse. On the one hand, this
interaction term explains the origin of the slow-down effect in the convergence
of Simultaneous Gradient Ascent (SGA) to stable Nash equilibria. On the other
hand, for the unstable equilibria, exponential convergence can be proved thanks
to the interaction term, for four modified dynamics proposed to stabilize GAN
training: Optimistic Mirror Descent (OMD), Consensus Optimization (CO),
Implicit Updates (IU) and Predictive Method (PM). The analysis uncovers the
intimate connections among these stabilizing techniques, and provides detailed
characterization on the choice of learning rate. As a by-product, we present a
new analysis for OMD proposed in Daskalakis, Ilyas, Syrgkanis, and Zeng [2017]
with improved rates. | [
"stat.ML",
"cs.GT",
"cs.LG"
] |
In this paper we present a deep learning method to estimate the illuminant of
an image. Our model is not trained with illuminant annotations, but with the
objective of improving performance on an auxiliary task such as object
recognition. To the best of our knowledge, this is the first example of a deep
learning architecture for illuminant estimation that is trained without ground
truth illuminants. We evaluate our solution on standard datasets for color
constancy, and compare it with state of the art methods. Our proposal is shown
to outperform most deep learning methods in a cross-dataset evaluation setup,
and to present competitive results in a comparison with parametric solutions. | [
"cs.CV"
] |
Knowledge Graph (KG) embedding has attracted more attention in recent years.
Most KG embedding models learn from time-unaware triples. However, the
inclusion of temporal information beside triples would further improve the
performance of a KGE model. In this regard, we propose ATiSE, a temporal KG
embedding model which incorporates time information into entity/relation
representations by using Additive Time Series decomposition. Moreover,
considering the temporal uncertainty during the evolution of entity/relation
representations over time, we map the representations of temporal KGs into the
space of multi-dimensional Gaussian distributions. The mean of each
entity/relation embedding at a time step shows the current expected position,
whereas its covariance (which is temporally stationary) represents its temporal
uncertainty. Experimental results show that ATiSE chieves the state-of-the-art
on link prediction over four temporal KGs. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Robust anomaly detection is a requirement for monitoring complex modern
systems with applications such as cyber-security, fraud prevention, and
maintenance. These systems generate multiple correlated time series that are
highly seasonal and noisy. This paper presents a novel unsupervised deep
learning architecture for multivariate time series anomaly detection, called
Robust Seasonal Multivariate Generative Adversarial Network (RSM-GAN). It
extends recent advancements in GANs with adoption of convolutional-LSTM layers
and an attention mechanism to produce state-of-the-art performance. We conduct
extensive experiments to demonstrate the strength of our architecture in
adjusting for complex seasonality patterns and handling severe levels of
training data contamination. We also propose a novel anomaly score assignment
and causal inference framework. We compare RSM-GAN with existing classical and
deep-learning based anomaly detection models, and the results show that our
architecture is associated with the lowest false positive rate and improves
precision by 30% and 16% in real-world and synthetic data, respectively.
Furthermore, we report the superiority of RSM-GAN regarding accurate root cause
identification and NAB scores in all data settings. | [
"cs.LG",
"stat.ML"
] |
This paper proposes a novel model for video generation and especially makes
the attempt to deal with the problem of video generation from text
descriptions, i.e., synthesizing realistic videos conditioned on given texts.
Existing video generation methods cannot be easily adapted to handle this task
well, due to the frame discontinuity issue and their text-free generation
schemes. To address these problems, we propose a recurrent deconvolutional
generative adversarial network (RD-GAN), which includes a recurrent
deconvolutional network (RDN) as the generator and a 3D convolutional neural
network (3D-CNN) as the discriminator. The RDN is a deconvolutional version of
conventional recurrent neural network, which can well model the long-range
temporal dependency of generated video frames and make good use of conditional
information. The proposed model can be jointly trained by pushing the RDN to
generate realistic videos so that the 3D-CNN cannot distinguish them from real
ones. We apply the proposed RD-GAN to a series of tasks including conventional
video generation, conditional video generation, video prediction and video
classification, and demonstrate its effectiveness by achieving well
performance. | [
"cs.CV"
] |
As digital medical imaging becomes more prevalent and archives increase in
size, representation learning exposes an interesting opportunity for enhanced
medical decision support systems. On the other hand, medical imaging data is
often scarce and short on annotations. In this paper, we present an assessment
of unsupervised feature learning approaches for images in the biomedical
literature, which can be applied to automatic biomedical concept detection. Six
unsupervised representation learning methods were built, including traditional
bags of visual words, autoencoders, and generative adversarial networks. Each
model was trained, and their respective feature space evaluated using images
from the ImageCLEF 2017 concept detection task. We conclude that it is possible
to obtain more powerful representations with modern deep learning approaches,
in contrast with previously popular computer vision methods. Although
generative adversarial networks can provide good results, they are harder to
succeed in highly varied data sets. The possibility of semi-supervised
learning, as well as their use in medical information retrieval problems, are
the next steps to be strongly considered. | [
"cs.CV"
] |
Invariant and equivariant networks have been successfully used for learning
images, sets, point clouds, and graphs. A basic challenge in developing such
networks is finding the maximal collection of invariant and equivariant linear
layers. Although this question is answered for the first three examples (for
popular transformations, at-least), a full characterization of invariant and
equivariant linear layers for graphs is not known.
In this paper we provide a characterization of all permutation invariant and
equivariant linear layers for (hyper-)graph data, and show that their
dimension, in case of edge-value graph data, is 2 and 15, respectively. More
generally, for graph data defined on k-tuples of nodes, the dimension is the
k-th and 2k-th Bell numbers. Orthogonal bases for the layers are computed,
including generalization to multi-graph data. The constant number of basis
elements and their characteristics allow successfully applying the networks to
different size graphs. From the theoretical point of view, our results
generalize and unify recent advancement in equivariant deep learning. In
particular, we show that our model is capable of approximating any message
passing neural network
Applying these new linear layers in a simple deep neural network framework is
shown to achieve comparable results to state-of-the-art and to have better
expressivity than previous invariant and equivariant bases. | [
"cs.LG",
"stat.ML"
] |
The target representation learned by convolutional neural networks plays an
important role in Thermal Infrared (TIR) tracking. Currently, most of the
top-performing TIR trackers are still employing representations learned by the
model trained on the RGB data. However, this representation does not take into
account the information in the TIR modality itself, limiting the performance of
TIR tracking. To solve this problem, we propose to distill representations of
the TIR modality from the RGB modality with Cross-Modal Distillation (CMD) on a
large amount of unlabeled paired RGB-TIR data. We take advantage of the
two-branch architecture of the baseline tracker, i.e. DiMP, for cross-modal
distillation working on two components of the tracker. Specifically, we use one
branch as a teacher module to distill the representation learned by the model
into the other branch. Benefiting from the powerful model in the RGB modality,
the cross-modal distillation can learn the TIR-specific representation for
promoting TIR tracking. The proposed approach can be incorporated into
different baseline trackers conveniently as a generic and independent
component. Furthermore, the semantic coherence of paired RGB and TIR images is
utilized as a supervised signal in the distillation loss for cross-modal
knowledge transfer. In practice, three different approaches are explored to
generate paired RGB-TIR patches with the same semantics for training in an
unsupervised way. It is easy to extend to an even larger scale of unlabeled
training data. Extensive experiments on the LSOTB-TIR dataset and PTB-TIR
dataset demonstrate that our proposed cross-modal distillation method
effectively learns TIR-specific target representations transferred from the RGB
modality. Our tracker outperforms the baseline tracker by achieving absolute
gains of 2.3% Success, 2.7% Precision, and 2.5% Normalized Precision
respectively. | [
"cs.CV"
] |
In real-world machine learning applications, data subsets correspond to
especially critical outcomes: vulnerable cyclist detections are safety-critical
in an autonomous driving task, and "question" sentences might be important to a
dialogue agent's language understanding for product purposes. While machine
learning models can achieve high quality performance on coarse-grained metrics
like F1-score and overall accuracy, they may underperform on critical
subsets---we define these as slices, the key abstraction in our approach. To
address slice-level performance, practitioners often train separate "expert"
models on slice subsets or use multi-task hard parameter sharing. We propose
Slice-based Learning, a new programming model in which the slicing function
(SF), a programming interface, specifies critical data subsets for which the
model should commit additional capacity. Any model can leverage SFs to learn
slice expert representations, which are combined with an attention mechanism to
make slice-aware predictions. We show that our approach maintains a
parameter-efficient representation while improving over baselines by up to 19.0
F1 on slices and 4.6 F1 overall on datasets spanning language understanding
(e.g. SuperGLUE), computer vision, and production-scale industrial systems. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
Every day, poison control centers (PCC) are called for immediate
classification and treatment recommendations if an acute intoxication is
suspected. Due to the time-sensitive nature of these cases, doctors are
required to propose a correct diagnosis and intervention within a minimal time
frame. Usually the toxin is known and recommendations can be made accordingly.
However, in challenging cases only symptoms are mentioned and doctors have to
rely on their clinical experience. Medical experts and our analyses of a
regional dataset of intoxication records provide evidence that this is
challenging, since occurring symptoms may not always match the textbook
description due to regional distinctions, inter-rater variance, and
institutional workflow. Computer-aided diagnosis (CADx) can provide decision
support, but approaches so far do not consider additional information of the
reported cases like age or gender, despite their potential value towards a
correct diagnosis. In this work, we propose a new machine learning based CADx
method which fuses symptoms and meta information of the patients using graph
convolutional networks. We further propose a novel symptom matching method that
allows the effective incorporation of prior knowledge into the learning process
and evidently stabilizes the poison prediction. We validate our method against
10 medical doctors with different experience diagnosing intoxication cases for
10 different toxins from the PCC in Munich and show our method's superiority in
performance for poison prediction. | [
"cs.LG",
"cs.AI",
"68T99"
] |
Knowledge is captured in the form of entities and their relationships and
stored in knowledge graphs. Knowledge graphs enhance the capabilities of
applications in many different areas including Web search, recommendation, and
natural language understanding. This is mainly because, entities enable
machines to understand things that go beyond simple tokens. Many modern
algorithms use learned entity embeddings from these structured representations.
However, building a knowledge graph takes time and effort, hence very costly
and nontrivial. On the other hand, many Web sources describe entities in some
structured format and therefore, finding ways to get them into useful entity
knowledge is advantageous. We propose an approach that processes entity centric
textual knowledge sources to learn entity embeddings and in turn avoids the
need for a traditional knowledge graph. We first extract triples into the new
representation format that does not use traditional complex triple extraction
methods defined by pre-determined relationship labels. Then we learn entity
embeddings through this new type of triples. We show that the embeddings
learned from our approach are: (i) high quality and comparable to a known
knowledge graph-based embeddings and can be used to improve them further, (ii)
better than a contextual language model-based entity embeddings, and (iii) easy
to compute and versatile in domain-specific applications where a knowledge
graph is not readily available | [
"cs.LG",
"cs.CL",
"cs.IR"
] |
We tackle the challenge of Visual Question Answering in multi-image setting
for the ISVQA dataset. Traditional VQA tasks have focused on a single-image
setting where the target answer is generated from a single image. Image set
VQA, however, comprises of a set of images and requires finding connection
between images, relate the objects across images based on these connections and
generate a unified answer. In this report, we work with 4 approaches in a bid
to improve the performance on the task. We analyse and compare our results with
three baseline models - LXMERT, HME-VideoQA and VisualBERT - and show that our
approaches can provide a slight improvement over the baselines. In specific, we
try to improve on the spatial awareness of the model and help the model
identify color using enhanced pre-training, reduce language dependence using
adversarial regularization, and improve counting using regression loss and
graph based deduplication. We further delve into an in-depth analysis on the
language bias in the ISVQA dataset and show how models trained on ISVQA
implicitly learn to associate language more strongly with the final answer. | [
"cs.CV",
"cs.CL",
"cs.LG"
] |
Analyzing motion between two consecutive images is one of the fundamental
tasks in computer vision. In the lack of labeled data, the loss functions are
split into consistency and smoothness, allowing for self-supervised training.
This paper focuses on the cost function derivation and presents an unrolling
iterative approach, transferring the hard L1 smoothness constraint into a
softer multi-layer iterative scheme. More accurate gradients, especially near
non-differential positions, improve the network's convergence, providing
superior results on tested scenarios. We report state-of-the-art results on
both MPI Sintel and KITTI 2015 unsupervised optical flow benchmarks. The
provided approach can be used to enhance various architectures and not limited
just to the presented pipeline. | [
"cs.CV",
"cs.LG"
] |
Synthesizing high quality saliency maps from noisy images is a challenging
problem in computer vision and has many practical applications. Samples
generated by existing techniques for saliency detection cannot handle the noise
perturbations smoothly and fail to delineate the salient objects present in the
given scene. In this paper, we present a novel end-to-end coupled Denoising
based Saliency Prediction with Generative Adversarial Network (DSAL-GAN)
framework to address the problem of salient object detection in noisy images.
DSAL-GAN consists of two generative adversarial-networks (GAN) trained
end-to-end to perform denoising and saliency prediction altogether in a
holistic manner. The first GAN consists of a generator which denoises the noisy
input image, and in the discriminator counterpart we check whether the output
is a denoised image or ground truth original image. The second GAN predicts the
saliency maps from raw pixels of the input denoised image using a data-driven
metric based on saliency prediction method with adversarial loss. Cycle
consistency loss is also incorporated to further improve salient region
prediction. We demonstrate with comprehensive evaluation that the proposed
framework outperforms several baseline saliency models on various performance
benchmarks. | [
"cs.CV"
] |
Zero-shot learning (ZSL) aims to discriminate images from unseen classes by
exploiting relations to seen classes via their attribute-based descriptions.
Since attributes are often related to specific parts of objects, many recent
works focus on discovering discriminative regions. However, these methods
usually require additional complex part detection modules or attention
mechanisms. In this paper, 1) we show that common ZSL backbones (without
explicit attention nor part detection) can implicitly localize attributes, yet
this property is not exploited. 2) Exploiting it, we then propose SELAR, a
simple method that further encourages attribute localization, surprisingly
achieving very competitive generalized ZSL (GZSL) performance when compared
with more complex state-of-the-art methods. Our findings provide useful insight
for designing future GZSL methods, and SELAR provides an easy to implement yet
strong baseline. | [
"cs.CV"
] |
Nowcasting is a field of meteorology which aims at forecasting weather on a
short term of up to a few hours. In the meteorology landscape, this field is
rather specific as it requires particular techniques, such as data
extrapolation, where conventional meteorology is generally based on physical
modeling. In this paper, we focus on cloud cover nowcasting, which has various
application areas such as satellite shots optimisation and photovoltaic energy
production forecast.
Following recent deep learning successes on multiple imagery tasks, we
applied deep convolutionnal neural networks on Meteosat satellite images for
cloud cover nowcasting. We present the results of several architectures
specialized in image segmentation and time series prediction. We selected the
best models according to machine learning metrics as well as meteorological
metrics. All selected architectures showed significant improvements over
persistence and the well-known U-Net surpasses AROME physical model. | [
"cs.CV",
"cs.AI",
"cs.LG"
] |
Generative Adversarial Networks (GANs) have recently achieved significant
improvement on paired/unpaired image-to-image translation, such as
photo$\rightarrow$ sketch and artist painting style transfer. However, existing
models can only be capable of transferring the low-level information (e.g.
color or texture changes), but fail to edit high-level semantic meanings (e.g.,
geometric structure or content) of objects. On the other hand, while some
researches can synthesize compelling real-world images given a class label or
caption, they cannot condition on arbitrary shapes or structures, which largely
limits their application scenarios and interpretive capability of model
results. In this work, we focus on a more challenging semantic manipulation
task, which aims to modify the semantic meaning of an object while preserving
its own characteristics (e.g. viewpoints and shapes), such as
cow$\rightarrow$sheep, motor$\rightarrow$ bicycle, cat$\rightarrow$dog. To
tackle such large semantic changes, we introduce a contrasting GAN
(contrast-GAN) with a novel adversarial contrasting objective. Instead of
directly making the synthesized samples close to target data as previous GANs
did, our adversarial contrasting objective optimizes over the distance
comparisons between samples, that is, enforcing the manipulated data be
semantically closer to the real data with target category than the input data.
Equipped with the new contrasting objective, a novel mask-conditional
contrast-GAN architecture is proposed to enable disentangle image background
with object semantic changes. Experiments on several semantic manipulation
tasks on ImageNet and MSCOCO dataset show considerable performance gain by our
contrast-GAN over other conditional GANs. Quantitative results further
demonstrate the superiority of our model on generating manipulated results with
high visual fidelity and reasonable object semantics. | [
"cs.CV"
] |
Underwater image enhancement is an important low-level computer vision task
for autonomous underwater vehicles and remotely operated vehicles to explore
and understand the underwater environments. Recently, deep convolutional neural
networks (CNNs) have been successfully used in many computer vision problems,
and so does underwater image enhancement. There are many deep-learning-based
methods with impressive performance for underwater image enhancement, but their
memory and model parameter costs are hindrances in practical application. To
address this issue, we propose a lightweight adaptive feature fusion network
(LAFFNet). The model is the encoder-decoder model with multiple adaptive
feature fusion (AAF) modules. AAF subsumes multiple branches with different
kernel sizes to generate multi-scale feature maps. Furthermore, channel
attention is used to merge these feature maps adaptively. Our method reduces
the number of parameters from 2.5M to 0.15M (around 94% reduction) but
outperforms state-of-the-art algorithms by extensive experiments. Furthermore,
we demonstrate our LAFFNet effectively improves high-level vision tasks like
salience object detection and single image depth estimation. | [
"cs.CV"
] |
Knowledge of the importance of input features towards decisions made by
machine-learning models is essential to increase our understanding of both the
models and the underlying data. Here, we present a new approach to estimating
feature importance with neural networks based on the idea of distributing the
features of interest among experts in an attentive mixture of experts (AME).
AMEs use attentive gating networks trained with a Granger-causal objective to
learn to jointly produce accurate predictions as well as estimates of feature
importance in a single model. Our experiments show (i) that the feature
importance estimates provided by AMEs compare favourably to those provided by
state-of-the-art methods, (ii) that AMEs are significantly faster at estimating
feature importance than existing methods, and (iii) that the associations
discovered by AMEs are consistent with those reported by domain experts. | [
"cs.LG",
"cs.AI",
"cs.NE"
] |
This paper proposes the Parallel WiSARD Object Tracker (PWOT), a new object
tracker based on the WiSARD weightless neural network that is robust against
quantization errors. Object tracking in video is an important and challenging
task in many applications. Difficulties can arise due to weather conditions,
target trajectory and appearance, occlusions, lighting conditions and noise.
Tracking is a high-level application and requires the object location frame by
frame in real time. This paper proposes a fast hybrid image segmentation
(threshold and edge detection) in YcbCr color model and a parallel RAM based
discriminator that improves efficiency when quantization errors occur. The
original WiSARD training algorithm was changed to allow the tracking. | [
"cs.CV"
] |
Few-shot object detection is a challenging but realistic scenario, where only
a few annotated training images are available for training detectors. A popular
approach to handle this problem is transfer learning, i.e., fine-tuning a
detector pretrained on a source-domain benchmark. However, such transferred
detector often fails to recognize new objects in the target domain, due to low
data diversity of training samples. To tackle this problem, we propose a novel
Context-Transformer within a concise deep transfer framework. Specifically,
Context-Transformer can effectively leverage source-domain object knowledge as
guidance, and automatically exploit contexts from only a few training images in
the target domain. Subsequently, it can adaptively integrate these relational
clues to enhance the discriminative power of detector, in order to reduce
object confusion in few-shot scenarios. Moreover, Context-Transformer is
flexibly embedded in the popular SSD-style detectors, which makes it a
plug-and-play module for end-to-end few-shot learning. Finally, we evaluate
Context-Transformer on the challenging settings of few-shot detection and
incremental few-shot detection. The experimental results show that, our
framework outperforms the recent state-of-the-art approaches. | [
"cs.CV",
"cs.LG"
] |
Many mobile systems and wearable devices, such as Virtual Reality (VR) or
Augmented Reality (AR) headsets, lack a keyboard or touchscreen to type an ID
and password for signing into a virtual website. However, they are usually
equipped with gesture capture interfaces to allow the user to interact with the
system directly with hand gestures. Although gesture-based authentication has
been well-studied, less attention is paid to the gesture-based user
identification problem, which is essentially an input method of account ID and
an efficient searching and indexing method of a database of gesture signals. In
this paper, we propose FMHash (i.e., Finger Motion Hash), a user identification
framework that can generate a compact binary hash code from a piece of
in-air-handwriting of an ID string. This hash code enables indexing and fast
search of a large account database using the in-air-handwriting by a hash
table. To demonstrate the effectiveness of the framework, we implemented a
prototype and achieved >99.5% precision and >92.6% recall with exact hash code
match on a dataset of 200 accounts collected by us. The ability of hashing
in-air-handwriting pattern to binary code can be used to achieve convenient
sign-in and sign-up with in-air-handwriting gesture ID on future mobile and
wearable systems connected to the Internet. | [
"cs.CV",
"D.4.6; I.5.4"
] |
Deep learning-based semi-supervised learning (SSL) algorithms have led to
promising results in medical images segmentation and can alleviate doctors'
expensive annotations by leveraging unlabeled data. However, most of the
existing SSL algorithms in literature tend to regularize the model training by
perturbing networks and/or data. Observing that multi/dual-task learning
attends to various levels of information which have inherent prediction
perturbation, we ask the question in this work: can we explicitly build
task-level regularization rather than implicitly constructing networks- and/or
data-level perturbation-and-transformation for SSL? To answer this question, we
propose a novel dual-task-consistency semi-supervised framework for the first
time. Concretely, we use a dual-task deep network that jointly predicts a
pixel-wise segmentation map and a geometry-aware level set representation of
the target. The level set representation is converted to an approximated
segmentation map through a differentiable task transform layer. Simultaneously,
we introduce a dual-task consistency regularization between the level
set-derived segmentation maps and directly predicted segmentation maps for both
labeled and unlabeled data. Extensive experiments on two public datasets show
that our method can largely improve the performance by incorporating the
unlabeled data. Meanwhile, our framework outperforms the state-of-the-art
semi-supervised medical image segmentation methods. Code is available at:
https://github.com/Luoxd1996/DTC | [
"cs.CV"
] |
Hashing method maps similar data to binary hashcodes with smaller hamming
distance, which has received a broad attention due to its low storage cost and
fast retrieval speed. With the rapid development of deep learning, deep hashing
methods have achieved promising results in efficient information retrieval.
Most of the existing deep hashing methods adopt pairwise or triplet losses to
deal with similarities underlying the data, but the training is difficult and
less efficient because $O(n^2)$ data pairs and $O(n^3)$ triplets are involved.
To address these issues, we propose a novel deep hashing algorithm with unary
loss which can be trained very efficiently. We first of all introduce a Unary
Upper Bound of the traditional triplet loss, thus reducing the complexity to
$O(n)$ and bridging the classification-based unary loss and the triplet loss.
Second, we propose a novel Semantic Cluster Deep Hashing (SCDH) algorithm by
introducing a modified Unary Upper Bound loss, named Semantic Cluster Unary
Loss (SCUL). The resultant hashcodes form several compact clusters, which means
hashcodes in the same cluster have similar semantic information. We also
demonstrate that the proposed SCDH is easy to be extended to semi-supervised
settings by incorporating the state-of-the-art semi-supervised learning
algorithms. Experiments on large-scale datasets show that the proposed method
is superior to state-of-the-art hashing algorithms. | [
"cs.CV"
] |
Thanks to their ability to learn flexible data-driven losses, Generative
Adversarial Networks (GANs) are an integral part of many semi- and
weakly-supervised methods for medical image segmentation. GANs jointly optimise
a generator and an adversarial discriminator on a set of training data. After
training has completed, the discriminator is usually discarded and only the
generator is used for inference. But should we discard discriminators? In this
work, we argue that training stable discriminators produces expressive loss
functions that we can re-use at inference to detect and correct segmentation
mistakes. First, we identify key challenges and suggest possible solutions to
make discriminators re-usable at inference. Then, we show that we can combine
discriminators with image reconstruction costs (via decoders) to further
improve the model. Our method is simple and improves the test-time performance
of pre-trained GANs. Moreover, we show that it is compatible with standard
post-processing techniques and it has potentials to be used for Online
Continual Learning. With our work, we open new research avenues for re-using
adversarial discriminators at inference. | [
"cs.CV",
"eess.IV"
] |
Studying competition and market structure at the product level instead of
brand level can provide firms with insights on cannibalization and product line
optimization. However, it is computationally challenging to analyze
product-level competition for the millions of products available on e-commerce
platforms. We introduce Product2Vec, a method based on the representation
learning algorithm Word2Vec, to study product-level competition, when the
number of products is large. The proposed model takes shopping baskets as
inputs and, for every product, generates a low-dimensional embedding that
preserves important product information. In order for the product embeddings to
be useful for firm strategic decision making, we leverage economic theories and
causal inference to propose two modifications to Word2Vec. First of all, we
create two measures, complementarity and exchangeability, that allow us to
determine whether product pairs are complements or substitutes. Second, we
combine these vectors with random utility-based choice models to forecast
demand. To accurately estimate price elasticities, i.e., how demand responds to
changes in price, we modify Word2Vec by removing the influence of price from
the product vectors. We show that, compared with state-of-the-art models, our
approach is faster, and can produce more accurate demand forecasts and price
elasticities. | [
"cs.LG",
"stat.ML"
] |
Over the past few years many research efforts have been devoted to the field
of affect analysis. Various approaches have been proposed for: i) discrete
emotion recognition in terms of the primary facial expressions; ii) emotion
analysis in terms of facial Action Units (AUs), assuming a fixed expression
intensity; iii) dimensional emotion analysis, in terms of valence and arousal
(VA). These approaches can only be effective, if they are developed using
large, appropriately annotated databases, showing behaviors of people
in-the-wild, i.e., in uncontrolled environments. Aff-Wild has been the first,
large-scale, in-the-wild database (including around 1,200,000 frames of 300
videos), annotated in terms of VA. In the vast majority of existing emotion
databases, their annotation is limited to either primary expressions, or
valence-arousal, or action units. In this paper, we first annotate a part
(around $234,000$ frames) of the Aff-Wild database in terms of $8$ AUs and
another part (around $288,000$ frames) in terms of the $7$ basic emotion
categories, so that parts of this database are annotated in terms of VA, as
well as AUs, or primary expressions. Then, we set up and tackle multi-task
learning for emotion recognition, as well as for facial image generation.
Multi-task learning is performed using: i) a deep neural network with shared
hidden layers, which learns emotional attributes by exploiting their
inter-dependencies; ii) a discriminator of a generative adversarial network
(GAN). On the other hand, image generation is implemented through the generator
of the GAN. For these two tasks, we carefully design loss functions that fit
the examined set-up. Experiments are presented which illustrate the good
performance of the proposed approach when applied to the new annotated parts of
the Aff-Wild database. | [
"cs.CV",
"cs.AI",
"cs.LG",
"stat.ML"
] |
Although transfer learning is proven to be effective in computer vision and
natural language processing applications, it is rarely investigated in
forecasting financial time series. Majority of existing works on transfer
learning are based on single-source transfer learning due to the availability
of open-access large-scale datasets. However, in financial domain, the lengths
of individual time series are relatively short and single-source transfer
learning models are less effective. Therefore, in this paper, we investigate
multi-source deep transfer learning for financial time series. We propose two
multi-source transfer learning methods namely Weighted Average Ensemble for
Transfer Learning (WAETL) and Tree-structured Parzen Estimator Ensemble
Selection (TPEES). The effectiveness of our approach is evaluated on financial
time series extracted from stock markets. Experiment results reveal that TPEES
outperforms other baseline methods on majority of multi-source transfer tasks. | [
"cs.LG"
] |
We propose an automatic method for generating high-quality annotations for
depth-based hand segmentation, and introduce a large-scale hand segmentation
dataset. Existing datasets are typically limited to a single hand. By
exploiting the visual cues given by an RGBD sensor and a pair of colored
gloves, we automatically generate dense annotations for two hand segmentation.
This lowers the cost/complexity of creating high quality datasets, and makes it
easy to expand the dataset in the future. We further show that existing
datasets, even with data augmentation, are not sufficient to train a hand
segmentation algorithm that can distinguish two hands. Source and datasets will
be made publicly available. | [
"cs.CV"
] |
Estimating and optimizing Mutual Information (MI) is core to many problems in
machine learning; however, bounding MI in high dimensions is challenging. To
establish tractable and scalable objectives, recent work has turned to
variational bounds parameterized by neural networks, but the relationships and
tradeoffs between these bounds remains unclear. In this work, we unify these
recent developments in a single framework. We find that the existing
variational lower bounds degrade when the MI is large, exhibiting either high
bias or high variance. To address this problem, we introduce a continuum of
lower bounds that encompasses previous bounds and flexibly trades off bias and
variance. On high-dimensional, controlled problems, we empirically characterize
the bias and variance of the bounds and their gradients and demonstrate the
effectiveness of our new bounds for estimation and representation learning. | [
"cs.LG",
"stat.ML"
] |
Humans are very good at directing their visual attention toward relevant
areas when they search for different types of objects. For instance, when we
search for cars, we will look at the streets, not at the top of buildings. The
motivation of this paper is to train a network to do the same via a multi-task
learning approach. To train visual attention, we produce foreground/background
segmentation labels in a semi-supervised way, using background subtraction or
optical flow. Using these labels, we train an object detection model to produce
foreground/background segmentation maps as well as bounding boxes while sharing
most model parameters. We use those segmentation maps inside the network as a
self-attention mechanism to weight the feature map used to produce the bounding
boxes, decreasing the signal of non-relevant areas. We show that by using this
method, we obtain a significant mAP improvement on two traffic surveillance
datasets, with state-of-the-art results on both UA-DETRAC and UAVDT. | [
"cs.CV"
] |
In many optimization problems in wireless communications, the expressions of
objective function or constraints are hard or even impossible to derive, which
makes the solutions difficult to find. In this paper, we propose a model-free
learning framework to solve constrained optimization problems without the
supervision of the optimal solution. Neural networks are used respectively for
parameterizing the function to be optimized, parameterizing the Lagrange
multiplier associated with instantaneous constraints, and approximating the
unknown objective function or constraints. We provide learning algorithms to
train all the neural networks simultaneously, and reveal the connections of the
proposed framework with reinforcement learning. Numerical and simulation
results validate the proposed framework and demonstrate the efficiency of
model-free learning by taking power control problem as an example. | [
"cs.LG",
"eess.SP",
"stat.ML"
] |
Contrastive learning, which aims at minimizing the distance between positive
pairs while maximizing that of negative ones, has been widely and successfully
applied in unsupervised feature learning, where the design of positive and
negative (pos/neg) pairs is one of its keys. In this paper, we attempt to
devise a feature-level data manipulation, differing from data augmentation, to
enhance the generic contrastive self-supervised learning. To this end, we first
design a visualization scheme for pos/neg score (Pos/neg score indicates cosine
similarity of pos/neg pair.) distribution, which enables us to analyze,
interpret and understand the learning process. To our knowledge, this is the
first attempt of its kind. More importantly, leveraging this tool, we gain some
significant observations, which inspire our novel Feature Transformation
proposals including the extrapolation of positives. This operation creates
harder positives to boost the learning because hard positives enable the model
to be more view-invariant. Besides, we propose the interpolation among
negatives, which provides diversified negatives and makes the model more
discriminative. It is the first attempt to deal with both challenges
simultaneously. Experiment results show that our proposed Feature
Transformation can improve at least 6.0% accuracy on ImageNet-100 over MoCo
baseline, and about 2.0% accuracy on ImageNet-1K over the MoCoV2 baseline.
Transferring to the downstream tasks successfully demonstrate our model is less
task-bias. Visualization tools and codes
https://github.com/DTennant/CL-Visualizing-Feature-Transformation . | [
"cs.CV"
] |
The policy gradients of the expected return objective can react slowly to
rare rewards. Yet, in some cases agents may wish to emphasize the low or high
returns regardless of their probability. Borrowing from the economics and
control literature, we review the risk-sensitive value function that arises
from an exponential utility and illustrate its effects on an example. This
risk-sensitive value function is not always applicable to reinforcement
learning problems, so we introduce the particle value function defined by a
particle filter over the distributions of an agent's experience, which bounds
the risk-sensitive one. We illustrate the benefit of the policy gradients of
this objective in Cliffworld. | [
"cs.LG",
"cs.AI"
] |
Policy gradient is an efficient technique for improving a policy in a
reinforcement learning setting. However, vanilla online variants are on-policy
only and not able to take advantage of off-policy data. In this paper we
describe a new technique that combines policy gradient with off-policy
Q-learning, drawing experience from a replay buffer. This is motivated by
making a connection between the fixed points of the regularized policy gradient
algorithm and the Q-values. This connection allows us to estimate the Q-values
from the action preferences of the policy, to which we apply Q-learning
updates. We refer to the new technique as 'PGQL', for policy gradient and
Q-learning. We also establish an equivalency between action-value fitting
techniques and actor-critic algorithms, showing that regularized policy
gradient techniques can be interpreted as advantage function learning
algorithms. We conclude with some numerical examples that demonstrate improved
data efficiency and stability of PGQL. In particular, we tested PGQL on the
full suite of Atari games and achieved performance exceeding that of both
asynchronous advantage actor-critic (A3C) and Q-learning. | [
"cs.LG",
"cs.AI",
"math.OC",
"stat.ML"
] |
Neuroimaging techniques have shown to be useful when studying the brain's
activity. This paper uses Magnetoencephalography (MEG) data, provided by the
Human Connectome Project (HCP), in combination with various deep artificial
neural network models to perform brain decoding. More specifically, here we
investigate to which extent can we infer the task performed by a subject based
on its MEG data. Three models based on compact convolution, combined
convolutional and long short-term architecture as well as a model based on
multi-view learning that aims at fusing the outputs of the two stream networks
are proposed and examined. These models exploit the spatio-temporal MEG data
for learning new representations that are used to decode the relevant tasks
across subjects. In order to realize the most relevant features of the input
signals, two attention mechanisms, i.e. self and global attention, are
incorporated in all the models. The experimental results of cross subject
multi-class classification on the studied MEG dataset show that the inclusion
of attention improves the generalization of the models across subjects. | [
"cs.LG",
"eess.SP",
"q-bio.NC",
"stat.ML",
"I.2; I.5"
] |
Glioblastoma is profoundly heterogeneous in regional microstructure and
vasculature. Characterizing the spatial heterogeneity of glioblastoma could
lead to more precise treatment. With unsupervised learning techniques,
glioblastoma MRI-derived radiomic features have been widely utilized for tumor
sub-region segmentation and survival prediction. However, the reliability of
algorithm outcomes is often challenged by both ambiguous intermediate process
and instability introduced by the randomness of clustering algorithms,
especially for data from heterogeneous patients.
In this paper, we propose an adaptive unsupervised learning approach for
efficient MRI intra-tumor partitioning and glioblastoma survival prediction. A
novel and problem-specific Feature-enhanced Auto-Encoder (FAE) is developed to
enhance the representation of pairwise clinical modalities and therefore
improve clustering stability of unsupervised learning algorithms such as
K-means. Moreover, the entire process is modelled by the Bayesian optimization
(BO) technique with a custom loss function that the hyper-parameters can be
adaptively optimized in a reasonably few steps. The results demonstrate that
the proposed approach can produce robust and clinically relevant MRI
sub-regions and statistically significant survival predictions. | [
"cs.LG",
"eess.SP"
] |
In complex networks, nodes that share similar structural characteristics
often exhibit similar roles (e.g type of users in a social network or the
hierarchical position of employees in a company). In order to leverage this
relationship, a growing literature proposed latent representations that
identify structurally equivalent nodes. However, most of the existing methods
require high time and space complexity. In this paper, we propose VNEstruct, a
simple approach for generating low-dimensional structural node embeddings, that
is both time efficient and robust to perturbations of the graph structure. The
proposed approach focuses on the local neighborhood of each node and employs
the Von Neumann entropy, an information-theoretic tool, to extract features
that capture the neighborhood's topology. Moreover, on graph classification
tasks, we suggest the utilization of the generated structural embeddings for
the transformation of an attributed graph structure into a set of augmented
node attributes. Empirically, we observe that the proposed approach exhibits
robustness on structural role identification tasks and state-of-the-art
performance on graph classification tasks, while maintaining very high
computational speed. | [
"cs.LG",
"cs.SI",
"stat.ML"
] |
We present XEM, an eXplainable Ensemble method for Multivariate time series
classification. XEM relies on a new hybrid ensemble method that combines an
explicit boosting-bagging approach to handle the bias-variance trade-off faced
by machine learning models and an implicit divide-and-conquer approach to
individualize classifier errors on different parts of the training data. Our
evaluation shows that XEM outperforms the state-of-the-art MTS classifiers on
the UEA datasets. Furthermore, XEM provides faithful explainability by design
and manifests robust performance when faced with challenges arising from
continuous data collection (different MTS length, missing data and noise). | [
"cs.LG",
"stat.ML"
] |
We present Deformable PV-RCNN, a high-performing point-cloud based 3D object
detector. Currently, the proposal refinement methods used by the
state-of-the-art two-stage detectors cannot adequately accommodate differing
object scales, varying point-cloud density, part-deformation and clutter. We
present a proposal refinement module inspired by 2D deformable convolution
networks that can adaptively gather instance-specific features from locations
where informative content exists. We also propose a simple context gating
mechanism which allows the keypoints to select relevant context information for
the refinement stage. We show state-of-the-art results on the KITTI dataset. | [
"cs.CV",
"cs.LG"
] |
In this paper, we focus on graph representation learning of heterogeneous
information network (HIN), in which various types of vertices are connected by
various types of relations. Most of the existing methods conducted on HIN
revise homogeneous graph embedding models via meta-paths to learn
low-dimensional vector space of HIN. In this paper, we propose a novel
Heterogeneous Graph Structural Attention Neural Network (HetSANN) to directly
encode structural information of HIN without meta-path and achieve more
informative representations. With this method, domain experts will not be
needed to design meta-path schemes and the heterogeneous information can be
processed automatically by our proposed model. Specifically, we implicitly
represent heterogeneous information using the following two methods: 1) we
model the transformation between heterogeneous vertices through a projection in
low-dimensional entity spaces; 2) afterwards, we apply the graph neural network
to aggregate multi-relational information of projected neighborhood by means of
attention mechanism. We also present three extensions of HetSANN, i.e.,
voices-sharing product attention for the pairwise relationships in HIN,
cycle-consistency loss to retain the transformation between heterogeneous
entity spaces, and multi-task learning with full use of information. The
experiments conducted on three public datasets demonstrate that our proposed
models achieve significant and consistent improvements compared to
state-of-the-art solutions. | [
"cs.LG",
"cs.SI",
"stat.ML"
] |
Despite remarkable empirical success, the training dynamics of generative
adversarial networks (GAN), which involves solving a minimax game using
stochastic gradients, is still poorly understood. In this work, we analyze
last-iterate convergence of simultaneous gradient descent (simGD) and its
variants under the assumption of convex-concavity, guided by a continuous-time
analysis with differential equations. First, we show that simGD, as is,
converges with stochastic sub-gradients under strict convexity in the primal
variable. Second, we generalize optimistic simGD to accommodate an optimism
rate separate from the learning rate and show its convergence with full
gradients. Finally, we present anchored simGD, a new method, and show
convergence with stochastic subgradients. | [
"cs.LG",
"stat.ML"
] |
We propose a new multi-frame method for efficiently computing scene flow
(dense depth and optical flow) and camera ego-motion for a dynamic scene
observed from a moving stereo camera rig. Our technique also segments out
moving objects from the rigid scene. In our method, we first estimate the
disparity map and the 6-DOF camera motion using stereo matching and visual
odometry. We then identify regions inconsistent with the estimated camera
motion and compute per-pixel optical flow only at these regions. This flow
proposal is fused with the camera motion-based flow proposal using fusion moves
to obtain the final optical flow and motion segmentation. This unified
framework benefits all four tasks - stereo, optical flow, visual odometry and
motion segmentation leading to overall higher accuracy and efficiency. Our
method is currently ranked third on the KITTI 2015 scene flow benchmark.
Furthermore, our CPU implementation runs in 2-3 seconds per frame which is 1-3
orders of magnitude faster than the top six methods. We also report a thorough
evaluation on challenging Sintel sequences with fast camera and object motion,
where our method consistently outperforms OSF [Menze and Geiger, 2015], which
is currently ranked second on the KITTI benchmark. | [
"cs.CV"
] |
Although Transformer has made breakthrough success in widespread domains
especially in Natural Language Processing (NLP), applying it to time series
forecasting is still a great challenge. In time series forecasting, the
autoregressive decoding of canonical Transformer models could introduce huge
accumulative errors inevitably. Besides, utilizing Transformer to deal with
spatial-temporal dependencies in the problem still faces tough difficulties.~To
tackle these limitations, this work is the first attempt to propose a
Non-Autoregressive Transformer architecture for time series forecasting, aiming
at overcoming the time delay and accumulative error issues in the canonical
Transformer. Moreover, we present a novel spatial-temporal attention mechanism,
building a bridge by a learned temporal influence map to fill the gaps between
the spatial and temporal attention, so that spatial and temporal dependencies
can be processed integrally. Empirically, we evaluate our model on diversified
ego-centric future localization datasets and demonstrate state-of-the-art
performance on both real-time and accuracy. | [
"cs.LG",
"stat.ML"
] |
When trained on multimodal image datasets, normal Generative Adversarial
Networks (GANs) are usually outperformed by class-conditional GANs and ensemble
GANs, but conditional GANs is restricted to labeled datasets and ensemble GANs
lack efficiency. We propose a novel GAN variant called virtual conditional GAN
(vcGAN) which is not only an ensemble GAN with multiple generative paths while
adding almost zero network parameters, but also a conditional GAN that can be
trained on unlabeled datasets without explicit clustering steps or objectives
other than the adversary loss. Inside the vcGAN's generator, a learnable
``analog-to-digital converter (ADC)" module maps a slice of the inputted
multivariate Gaussian noise to discrete/digital noise (virtual label),
according to which a selector selects the corresponding generative path to
produce the sample. All the generative paths share the same decoder network
while in each path the decoder network is fed with a concatenation of a
different pre-computed amplified one-hot vector and the inputted Gaussian
noise. We conducted a lot of experiments on several balanced/imbalanced image
datasets to demonstrate that vcGAN converges faster and achieves improved
Frech\'et Inception Distance (FID). In addition, we show the training byproduct
that the ADC in vcGAN learned the categorical probability of each mode and that
each generative path generates samples of specific mode, which enables
class-conditional sampling. Codes are available at
\url{https://github.com/annonnymmouss/vcgan} | [
"cs.CV"
] |
Generative Adversarial Networks (GAN) receive great attentions recently due
to its excellent performance in image generation, transformation, and
super-resolution. However, GAN has rarely been studied and trained for
classification, leading that the generated images may not be appropriate for
classification. In this paper, we propose a novel Generative Adversarial
Classifier (GAC) particularly for low-resolution Handwriting Character
Recognition. Specifically, involving additionally a classifier in the training
process of normal GANs, GAC is calibrated for learning suitable structures and
restored characters images that benefits the classification. Experimental
results show that our proposed method can achieve remarkable performance in
handwriting characters 8x super-resolution, approximately 10% and 20% higher
than the present state-of-the-art methods respectively on benchmark data
CASIA-HWDB1.1 and MNIST. | [
"cs.CV",
"cs.AI",
"cs.LG"
] |
Deepfakes have become a critical social problem, and detecting them is of
utmost importance. Also, deepfake generation methods are advancing, and it is
becoming harder to detect. While many deepfake detection models can detect
different types of deepfakes separately, they perform poorly on generalizing
the detection performance over multiple types of deepfake. This motivates us to
develop a generalized model to detect different types of deepfakes. Therefore,
in this work, we introduce a practical digital forensic tool to detect
different types of deepfakes simultaneously and propose Transfer learning-based
Autoencoder with Residuals (TAR). The ultimate goal of our work is to develop a
unified model to detect various types of deepfake videos with high accuracy,
with only a small number of training samples that can work well in real-world
settings. We develop an autoencoder-based detection model with Residual blocks
and sequentially perform transfer learning to detect different types of
deepfakes simultaneously. Our approach achieves a much higher generalized
detection performance than the state-of-the-art methods on the FaceForensics++
dataset. In addition, we evaluate our model on 200 real-world
Deepfake-in-the-Wild (DW) videos of 50 celebrities available on the Internet
and achieve 89.49% zero-shot accuracy, which is significantly higher than the
best baseline model (gaining 10.77%), demonstrating and validating the
practicability of our approach. | [
"cs.CV"
] |
Performance monitoring of object detection is crucial for safety-critical
applications such as autonomous vehicles that operate under varying and complex
environmental conditions. Currently, object detectors are evaluated using
summary metrics based on a single dataset that is assumed to be representative
of all future deployment conditions. In practice, this assumption does not
hold, and the performance fluctuates as a function of the deployment
conditions. To address this issue, we propose an introspection approach to
performance monitoring during deployment without the need for ground truth
data. We do so by predicting when the per-frame mean average precision drops
below a critical threshold using the detector's internal features. We
quantitatively evaluate and demonstrate our method's ability to reduce risk by
trading off making an incorrect decision by raising the alarm and absenting
from detection. | [
"cs.CV"
] |
The goal in label-imbalanced and group-sensitive classification is to
optimize relevant metrics such as balanced error and equal opportunity.
Classical methods, such as weighted cross-entropy, fail when used with the
modern practice of training deep nets to the terminal phase of training(TPT),
that is training beyond zero training error. This observation has motivated
recent flurry of activity in developing heuristic alternatives following the
intuitive mechanism of promoting larger margin for minorities. In contrast to
previous heuristics, we follow a principled analysis explaining how different
loss adjustments affect margins. First, we prove that for all linear
classifiers trained in TPT, it is necessary to introduce multiplicative, rather
than additive, logit adjustments so that the relative margins between classes
change appropriately. To show this, we discover a connection of the
multiplicative CE modification to the so-called cost-sensitive support-vector
machines. Perhaps counterintuitively, we also find that, at the start of the
training, the same multiplicative weights can actually harm the minority
classes. Thus, while additive adjustments are ineffective in the TPT, we show
numerically that they can speed up convergence by countering the initial
negative effect of the multiplicative weights. Motivated by these findings, we
formulate the vector-scaling(VS) loss, that captures existing techniques as
special cases. Moreover, we introduce a natural extension of the VS-loss to
group-sensitive classification, thus treating the two common types of
imbalances (label/group) in a unifying way. Importantly, our experiments on
state-of-the-art datasets are fully consistent with our theoretical insights
and confirm the superior performance of our algorithms. Finally, for imbalanced
Gaussian-mixtures data, we perform a generalization analysis, revealing
tradeoffs between different metrics. | [
"cs.LG",
"stat.ML"
] |
Cross-modal retrieval between videos and texts has attracted growing
attentions due to the rapid emergence of videos on the web. The current
dominant approach for this problem is to learn a joint embedding space to
measure cross-modal similarities. However, simple joint embeddings are
insufficient to represent complicated visual and textual details, such as
scenes, objects, actions and their compositions. To improve fine-grained
video-text retrieval, we propose a Hierarchical Graph Reasoning (HGR) model,
which decomposes video-text matching into global-to-local levels. To be
specific, the model disentangles texts into hierarchical semantic graph
including three levels of events, actions, entities and relationships across
levels. Attention-based graph reasoning is utilized to generate hierarchical
textual embeddings, which can guide the learning of diverse and hierarchical
video representations. The HGR model aggregates matchings from different
video-text levels to capture both global and local details. Experimental
results on three video-text datasets demonstrate the advantages of our model.
Such hierarchical decomposition also enables better generalization across
datasets and improves the ability to distinguish fine-grained semantic
differences. | [
"cs.CV",
"cs.AI"
] |
Achieving transparency in black-box deep learning algorithms is still an open
challenge. High dimensional features and decisions given by deep neural
networks (NN) require new algorithms and methods to expose its mechanisms.
Current state-of-the-art NN interpretation methods (e.g. Saliency maps,
DeepLIFT, LIME, etc.) focus more on the direct relationship between NN outputs
and inputs rather than the NN structure and operations itself. In current deep
NN operations, there is uncertainty over the exact role played by neurons with
fixed activation functions. In this paper, we achieve partially explainable
learning model by symbolically explaining the role of activation functions (AF)
under a scalable topology. This is carried out by modeling the AFs as adaptive
Gaussian Processes (GP), which sit within a novel scalable NN topology, based
on the Kolmogorov-Arnold Superposition Theorem (KST). In this scalable NN
architecture, the AFs are generated by GP interpolation between control points
and can thus be tuned during the back-propagation procedure via gradient
descent. The control points act as the core enabler to both local and global
adjustability of AF, where the GP interpolation constrains the intrinsic
autocorrelation to avoid over-fitting. We show that there exists a trade-off
between the NN's expressive power and interpretation complexity, under linear
KST topology scaling. To demonstrate this, we perform a case study on a binary
classification dataset of banknote authentication. By quantitatively and
qualitatively investigating the mapping relationship between inputs and output,
our explainable model can provide interpretation over each of the
one-dimensional attributes. These early results suggest that our model has the
potential to act as the final interpretation layer for deep neural networks. | [
"cs.LG",
"cs.AI",
"stat.ML"
] |
One of the methods used in image recognition is the Deep Convolutional Neural
Network (DCNN). DCNN is a model in which the expressive power of features is
greatly improved by deepening the hidden layer of CNN. The architecture of CNNs
is determined based on a model of the visual cortex of mammals. There is a
model called Residual Network (ResNet) that has a skip connection. ResNet is an
advanced model in terms of the learning method, but it has not been interpreted
from a biological viewpoint. In this research, we investigate the receptive
fields of a ResNet on the classification task in ImageNet. We find that ResNet
has orientation selective neurons and double opponent color neurons. In
addition, we suggest that some inactive neurons in the first layer of ResNet
affect the classification task. | [
"cs.CV"
] |
We propose a novel capsule network based variational encoder architecture,
called Bayesian capsules (B-Caps), to modulate the mean and standard deviation
of the sampling distribution in the latent space. We hypothesized that this
approach can learn a better representation of features in the latent space than
traditional approaches. Our hypothesis was tested by using the learned latent
variables for image reconstruction task, where for MNIST and Fashion-MNIST
datasets, different classes were separated successfully in the latent space
using our proposed model. Our experimental results have shown improved
reconstruction and classification performances for both datasets adding
credence to our hypothesis. We also showed that by increasing the latent space
dimension, the proposed B-Caps was able to learn a better representation when
compared to the traditional variational auto-encoders (VAE). Hence our results
indicate the strength of capsule networks in representation learning which has
never been examined under the VAE settings before. | [
"cs.CV"
] |
Scientific workloads have traditionally exploited high levels of sparsity to
accelerate computation and reduce memory requirements. While deep neural
networks can be made sparse, achieving practical speedups on GPUs is difficult
because these applications have relatively moderate levels of sparsity that are
not sufficient for existing sparse kernels to outperform their dense
counterparts. In this work, we study sparse matrices from deep learning
applications and identify favorable properties that can be exploited to
accelerate computation. Based on these insights, we develop high-performance
GPU kernels for two sparse matrix operations widely applicable in neural
networks: sparse matrix-dense matrix multiplication and sampled dense-dense
matrix multiplication. Our kernels reach 27% of single-precision peak on Nvidia
V100 GPUs. Using our kernels, we demonstrate sparse Transformer and MobileNet
models that achieve 1.2-2.1x speedups and up to 12.8x memory savings without
sacrificing accuracy. | [
"cs.LG",
"cs.DC",
"stat.ML"
] |
Purpose: Lung nodules have very diverse shapes and sizes, which makes
classifying them as benign/malignant a challenging problem. In this paper, we
propose a novel method to predict the malignancy of nodules that have the
capability to analyze the shape and size of a nodule using a global feature
extractor, as well as the density and structure of the nodule using a local
feature extractor. Methods: We propose to use Residual Blocks with a 3x3 kernel
size for local feature extraction, and Non-Local Blocks to extract the global
features. The Non-Local Block has the ability to extract global features
without using a huge number of parameters. The key idea behind the Non-Local
Block is to apply matrix multiplications between features on the same feature
maps. Results: We trained and validated the proposed method on the LIDC-IDRI
dataset which contains 1,018 computed tomography (CT) scans. We followed a
rigorous procedure for experimental setup namely, 10-fold cross-validation and
ignored the nodules that had been annotated by less than 3 radiologists. The
proposed method achieved state-of-the-art results with AUC=95.62%, while
significantly outperforming other baseline methods. Conclusions: Our proposed
Deep Local-Global network has the capability to accurately extract both local
and global features. Our new method outperforms state-of-the-art architecture
including Densenet and Resnet with transfer learning. | [
"cs.CV",
"cs.AI",
"stat.ML"
] |
Generative adversarial networks (GANs) have achieved rapid progress in
learning rich data distributions. However, we argue about two main issues in
existing techniques. First, the low quality problem where the learned
distribution has massive low quality samples. Second, the missing modes problem
where the learned distribution misses some certain regions of the real data
distribution. To address these two issues, we propose a novel prior that
captures the whole real data distribution for GANs, which are called PriorGANs.
To be specific, we adopt a simple yet elegant Gaussian Mixture Model (GMM) to
build an explicit probability distribution on the feature level for the whole
real data. By maximizing the probability of generated data, we can push the low
quality samples to high quality. Meanwhile, equipped with the prior, we can
estimate the missing modes in the learned distribution and design a sampling
strategy on the real data to solve the problem. The proposed real data prior
can generalize to various training settings of GANs, such as LSGAN, WGAN-GP,
SNGAN, and even the StyleGAN. Our experiments demonstrate that PriorGANs
outperform the state-of-the-art on the CIFAR-10, FFHQ, LSUN-cat, and LSUN-bird
datasets by large margins. | [
"cs.CV",
"eess.IV"
] |
One intriguing property of deep neural networks (DNNs) is their inherent
vulnerability to backdoor attacks -- a trojan model responds to
trigger-embedded inputs in a highly predictable manner while functioning
normally otherwise. Despite the plethora of prior work on DNNs for continuous
data (e.g., images), the vulnerability of graph neural networks (GNNs) for
discrete-structured data (e.g., graphs) is largely unexplored, which is highly
concerning given their increasing use in security-sensitive domains. To bridge
this gap, we present GTA, the first backdoor attack on GNNs. Compared with
prior work, GTA departs in significant ways: graph-oriented -- it defines
triggers as specific subgraphs, including both topological structures and
descriptive features, entailing a large design spectrum for the adversary;
input-tailored -- it dynamically adapts triggers to individual graphs, thereby
optimizing both attack effectiveness and evasiveness; downstream model-agnostic
-- it can be readily launched without knowledge regarding downstream models or
fine-tuning strategies; and attack-extensible -- it can be instantiated for
both transductive (e.g., node classification) and inductive (e.g., graph
classification) tasks, constituting severe threats for a range of
security-critical applications. Through extensive evaluation using benchmark
datasets and state-of-the-art models, we demonstrate the effectiveness of GTA.
We further provide analytical justification for its effectiveness and discuss
potential countermeasures, pointing to several promising research directions. | [
"cs.LG",
"cs.CR",
"stat.ML"
] |
Referring expression comprehension (REF) aims at identifying a particular
object in a scene by a natural language expression. It requires joint reasoning
over the textual and visual domains to solve the problem. Some popular
referring expression datasets, however, fail to provide an ideal test bed for
evaluating the reasoning ability of the models, mainly because 1) their
expressions typically describe only some simple distinctive properties of the
object and 2) their images contain limited distracting information. To bridge
the gap, we propose a new dataset for visual reasoning in context of referring
expression comprehension with two main features. First, we design a novel
expression engine rendering various reasoning logics that can be flexibly
combined with rich visual properties to generate expressions with varying
compositionality. Second, to better exploit the full reasoning chain embodied
in an expression, we propose a new test setting by adding additional
distracting images containing objects sharing similar properties with the
referent, thus minimising the success rate of reasoning-free cross-domain
alignment. We evaluate several state-of-the-art REF models, but find none of
them can achieve promising performance. A proposed modular hard mining strategy
performs the best but still leaves substantial room for improvement. We hope
this new dataset and task can serve as a benchmark for deeper visual reasoning
analysis and foster the research on referring expression comprehension. | [
"cs.CV"
] |
There has been a strong push recently to examine biological scale simulations
of neuromorphic algorithms to achieve stronger inference capabilities. This
paper presents a set of piecewise linear spiking neuron models, which can
reproduce different behaviors, similar to the biological neuron, both for a
single neuron as well as a network of neurons. The proposed models are
investigated, in terms of digital implementation feasibility and costs,
targeting large scale hardware implementation. Hardware synthesis and physical
implementations on FPGA show that the proposed models can produce precise
neural behaviors with higher performance and considerably lower implementation
costs compared with the original model. Accordingly, a compact structure of the
models which can be trained with supervised and unsupervised learning
algorithms has been developed. Using this structure and based on a spike rate
coding, a character recognition case study has been implemented and tested. | [
"cs.LG",
"cs.NE",
"q-bio.NC"
] |
Although cameras are ubiquitous, robotic platforms typically rely on active
sensors like LiDAR for direct 3D perception. In this work, we propose a novel
self-supervised monocular depth estimation method combining geometry with a new
deep network, PackNet, learned only from unlabeled monocular videos. Our
architecture leverages novel symmetrical packing and unpacking blocks to
jointly learn to compress and decompress detail-preserving representations
using 3D convolutions. Although self-supervised, our method outperforms other
self, semi, and fully supervised methods on the KITTI benchmark. The 3D
inductive bias in PackNet enables it to scale with input resolution and number
of parameters without overfitting, generalizing better on out-of-domain data
such as the NuScenes dataset. Furthermore, it does not require large-scale
supervised pretraining on ImageNet and can run in real-time. Finally, we
release DDAD (Dense Depth for Automated Driving), a new urban driving dataset
with more challenging and accurate depth evaluation, thanks to longer-range and
denser ground-truth depth generated from high-density LiDARs mounted on a fleet
of self-driving cars operating world-wide. | [
"cs.CV",
"cs.LG",
"cs.RO"
] |
Deep learning methods for graphs have seen rapid progress in recent years
with much focus awarded to generalising Convolutional Neural Networks (CNN) to
graph data. CNNs are typically realised by alternating convolutional and
pooling layers where the pooling layers subsample the grid and exchange spatial
or temporal resolution for increased feature dimensionality. Whereas the
generalised convolution operator for graphs has been studied extensively and
proven useful, hierarchical coarsening of graphs is still challenging since
nodes in graphs have no spatial locality and no natural order. This paper
proposes two main contributions, the first is a differential module calculating
structural similarity features based on the adjacency matrix. These structural
similarity features may be used with various algorithms however in this paper
the focus and the second main contribution is on integrating these features
with a revisited pooling layer DiffPool arXiv:1806.08804 to propose a pooling
layer referred to as SimPool. This is achieved by linking the concept of
network reduction by means of structural similarity in graphs with the concept
of hierarchical localised pooling. Experimental results demonstrate that as
part of an end-to-end Graph Neural Network architecture SimPool calculates node
cluster assignments that functionally resemble more to the locality preserving
pooling operations used by CNNs that operate on local receptive fields in the
standard grid. Furthermore the experimental results demonstrate that these
features are useful in inductive graph classification tasks with no increase to
the number of parameters. | [
"cs.LG",
"stat.ML"
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.