April 3, 2020

2829 words 14 mins read

Paper Group ANR 15

Paper Group ANR 15

FragNet: Writer Identification using Deep Fragment Networks. Learning to Correct 3D Reconstructions from Multiple Views. Nature-Inspired Optimization Algorithms: Challenges and Open Problems. DeFeat-Net: General Monocular Depth via Simultaneous Unsupervised Representation Learning. Watching the World Go By: Representation Learning from Unlabeled Vi …

FragNet: Writer Identification using Deep Fragment Networks

Title FragNet: Writer Identification using Deep Fragment Networks
Authors Sheng He, Lambert Schomaker
Abstract Writer identification based on a small amount of text is a challenging problem. In this paper, we propose a new benchmark study for writer identification based on word or text block images which approximately contain one word. In order to extract powerful features on these word images, a deep neural network, named FragNet, is proposed. The FragNet has two pathways: feature pyramid which is used to extract feature maps and fragment pathway which is trained to predict the writer identity based on fragments extracted from the input image and the feature maps on the feature pyramid. We conduct experiments on four benchmark datasets, which show that our proposed method can generate efficient and robust deep representations for writer identification based on both word and page images.
Published 2020-03-16
URL https://arxiv.org/abs/2003.07212v2
PDF https://arxiv.org/pdf/2003.07212v2.pdf
PWC https://paperswithcode.com/paper/fragnet-writer-identification-using-deep

Learning to Correct 3D Reconstructions from Multiple Views

Title Learning to Correct 3D Reconstructions from Multiple Views
Authors Ştefan Săftescu, Paul Newman
Abstract This paper is about reducing the cost of building good large-scale 3D reconstructions post-hoc. We render 2D views of an existing reconstruction and train a convolutional neural network (CNN) that refines inverse-depth to match a higher-quality reconstruction. Since the views that we correct are rendered from the same reconstruction, they share the same geometry, so overlapping views complement each other. We take advantage of that in two ways. Firstly, we impose a loss during training which guides predictions on neighbouring views to have the same geometry and has been shown to improve performance. Secondly, in contrast to previous work, which corrects each view independently, we also make predictions on sets of neighbouring views jointly. This is achieved by warping feature maps between views and thus bypassing memory-intensive 3D computation. We make the observation that features in the feature maps are viewpoint-dependent, and propose a method for transforming features with dynamic filters generated by a multi-layer perceptron from the relative poses between views. In our experiments we show that this last step is necessary for successfully fusing feature maps between views.
Published 2020-01-22
URL https://arxiv.org/abs/2001.08098v1
PDF https://arxiv.org/pdf/2001.08098v1.pdf
PWC https://paperswithcode.com/paper/learning-to-correct-3d-reconstructions-from

Nature-Inspired Optimization Algorithms: Challenges and Open Problems

Title Nature-Inspired Optimization Algorithms: Challenges and Open Problems
Authors Xin-She Yang
Abstract Many problems in science and engineering can be formulated as optimization problems, subject to complex nonlinear constraints. The solutions of highly nonlinear problems usually require sophisticated optimization algorithms, and traditional algorithms may struggle to deal with such problems. A current trend is to use nature-inspired algorithms due to their flexibility and effectiveness. However, there are some key issues concerning nature-inspired computation and swarm intelligence. This paper provides an in-depth review of some recent nature-inspired algorithms with the emphasis on their search mechanisms and mathematical foundations. Some challenging issues are identified and five open problems are highlighted, concerning the analysis of algorithmic convergence and stability, parameter tuning, mathematical framework, role of benchmarking and scalability. These problems are discussed with the directions for future research.
Published 2020-03-08
URL https://arxiv.org/abs/2003.03776v1
PDF https://arxiv.org/pdf/2003.03776v1.pdf
PWC https://paperswithcode.com/paper/nature-inspired-optimization-algorithms

DeFeat-Net: General Monocular Depth via Simultaneous Unsupervised Representation Learning

Title DeFeat-Net: General Monocular Depth via Simultaneous Unsupervised Representation Learning
Authors Jaime Spencer, Richard Bowden, Simon Hadfield
Abstract In the current monocular depth research, the dominant approach is to employ unsupervised training on large datasets, driven by warped photometric consistency. Such approaches lack robustness and are unable to generalize to challenging domains such as nighttime scenes or adverse weather conditions where assumptions about photometric consistency break down. We propose DeFeat-Net (Depth & Feature network), an approach to simultaneously learn a cross-domain dense feature representation, alongside a robust depth-estimation framework based on warped feature consistency. The resulting feature representation is learned in an unsupervised manner with no explicit ground-truth correspondences required. We show that within a single domain, our technique is comparable to both the current state of the art in monocular depth estimation and supervised feature representation learning. However, by simultaneously learning features, depth and motion, our technique is able to generalize to challenging domains, allowing DeFeat-Net to outperform the current state-of-the-art with around 10% reduction in all error measures on more challenging sequences such as nighttime driving.
Tasks Depth Estimation, Monocular Depth Estimation, Representation Learning, Unsupervised Representation Learning
Published 2020-03-30
URL https://arxiv.org/abs/2003.13446v1
PDF https://arxiv.org/pdf/2003.13446v1.pdf
PWC https://paperswithcode.com/paper/defeat-net-general-monocular-depth-via

Watching the World Go By: Representation Learning from Unlabeled Videos

Title Watching the World Go By: Representation Learning from Unlabeled Videos
Authors Daniel Gordon, Kiana Ehsani, Dieter Fox, Ali Farhadi
Abstract Recent single image unsupervised representation learning techniques show remarkable success on a variety of tasks. The basic principle in these works is instance discrimination: learning to differentiate between two augmented versions of the same image and a large batch of unrelated images. Networks learn to ignore the augmentation noise and extract semantically meaningful representations. Prior work uses artificial data augmentation techniques such as cropping, and color jitter which can only affect the image in superficial ways and are not aligned with how objects actually change e.g. occlusion, deformation, viewpoint change. In this paper, we argue that videos offer this natural augmentation for free. Videos can provide entirely new views of objects, show deformation, and even connect semantically similar but visually distinct concepts. We propose Video Noise Contrastive Estimation, a method for using unlabeled video to learn strong, transferable single image representations. We demonstrate improvements over recent unsupervised single image techniques, as well as over fully supervised ImageNet pretraining, across a variety of temporal and non-temporal tasks.
Tasks Data Augmentation, Representation Learning, Unsupervised Representation Learning
Published 2020-03-18
URL https://arxiv.org/abs/2003.07990v1
PDF https://arxiv.org/pdf/2003.07990v1.pdf
PWC https://paperswithcode.com/paper/watching-the-world-go-by-representation

Controlling Computation versus Quality for Neural Sequence Models

Title Controlling Computation versus Quality for Neural Sequence Models
Authors Ankur Bapna, Naveen Arivazhagan, Orhan Firat
Abstract Most neural networks utilize the same amount of compute for every example independent of the inherent complexity of the input. Further, methods that adapt the amount of computation to the example focus on finding a fixed inference-time computational graph per example, ignoring any external computational budgets or varying inference time limitations. In this work, we utilize conditional computation to make neural sequence models (Transformer) more efficient and computation-aware during inference. We first modify the Transformer architecture, making each set of operations conditionally executable depending on the output of a learned control network. We then train this model in a multi-task setting, where each task corresponds to a particular computation budget. This allows us to train a single model that can be controlled to operate on different points of the computation-quality trade-off curve, depending on the available computation budget at inference time. We evaluate our approach on two tasks: (i) WMT English-French Translation and (ii) Unsupervised representation learning (BERT). Our experiments demonstrate that the proposed Conditional Computation Transformer (CCT) is competitive with vanilla Transformers when allowed to utilize its full computational budget, while improving significantly over computationally equivalent baselines when operating on smaller computational budgets.
Tasks Representation Learning, Unsupervised Representation Learning
Published 2020-02-17
URL https://arxiv.org/abs/2002.07106v1
PDF https://arxiv.org/pdf/2002.07106v1.pdf
PWC https://paperswithcode.com/paper/controlling-computation-versus-quality-for

Deep Self-Supervised Representation Learning for Free-Hand Sketch

Title Deep Self-Supervised Representation Learning for Free-Hand Sketch
Authors Peng Xu, Zeyu Song, Qiyue Yin, Yi-Zhe Song, Liang Wang
Abstract In this paper, we tackle for the first time, the problem of self-supervised representation learning for free-hand sketches. This importantly addresses a common problem faced by the sketch community – that annotated supervisory data are difficult to obtain. This problem is very challenging in that sketches are highly abstract and subject to different drawing styles, making existing solutions tailored for photos unsuitable. Key for the success of our self-supervised learning paradigm lies with our sketch-specific designs: (i) we propose a set of pretext tasks specifically designed for sketches that mimic different drawing styles, and (ii) we further exploit the use of a textual convolution network (TCN) in a dual-branch architecture for sketch feature learning, as means to accommodate the sequential stroke nature of sketches. We demonstrate the superiority of our sketch-specific designs through two sketch-related applications (retrieval and recognition) on a million-scale sketch dataset, and show that the proposed approach outperforms the state-of-the-art unsupervised representation learning methods, and significantly narrows the performance gap between with supervised representation learning.
Tasks Representation Learning, Unsupervised Representation Learning
Published 2020-02-03
URL https://arxiv.org/abs/2002.00867v1
PDF https://arxiv.org/pdf/2002.00867v1.pdf
PWC https://paperswithcode.com/paper/deep-self-supervised-representation-learning

Switching dynamics of single and coupled VO2-based oscillators as elements of neural networks

Title Switching dynamics of single and coupled VO2-based oscillators as elements of neural networks
Authors Andrei Velichko, Maksim Belyaev, Vadim Putrolaynen, Alexander Pergament, Valentin Perminov
Abstract In the present paper, we report on the switching dynamics of both single and coupled VO2-based oscillators, with resistive and capacitive coupling, and explore the capability of their application in oscillatory neural networks. Based on these results, we further select an adequate SPICE model to describe the modes of operation of coupled oscillator circuits. Physical mechanisms influencing the time of forward and reverse electrical switching, that determine the applicability limits of the proposed model, are identified. For the resistive coupling, it is shown that synchronization takes place at a certain value of the coupling resistance, though it is unstable and a synchronization failure occurs periodically. For the capacitive coupling, two synchronization modes, with weak and strong coupling, are found. The transition between these modes is accompanied by chaotic oscillations. A decrease in the width of the spectrum harmonics in the weak-coupling mode, and its increase in the strong-coupling one, is detected. The dependences of frequencies and phase differences of the coupled oscillatory circuits on the coupling capacitance are found. Examples of operation of coupled VO2 oscillators as a central pattern generator are demonstrated.
Published 2020-01-07
URL https://arxiv.org/abs/2001.01854v1
PDF https://arxiv.org/pdf/2001.01854v1.pdf
PWC https://paperswithcode.com/paper/switching-dynamics-of-single-and-coupled-vo2

Uncertainty Quantification for Sparse Deep Learning

Title Uncertainty Quantification for Sparse Deep Learning
Authors Yuexi Wang, Veronika Ročková
Abstract Deep learning methods continue to have a decided impact on machine learning, both in theory and in practice. Statistical theoretical developments have been mostly concerned with approximability or rates of estimation when recovering infinite dimensional objects (curves or densities). Despite the impressive array of available theoretical results, the literature has been largely silent about uncertainty quantification for deep learning. This paper takes a step forward in this important direction by taking a Bayesian point of view. We study Gaussian approximability of certain aspects of posterior distributions of sparse deep ReLU architectures in non-parametric regression. Building on tools from Bayesian non-parametrics, we provide semi-parametric Bernstein-von Mises theorems for linear and quadratic functionals, which guarantee that implied Bayesian credible regions have valid frequentist coverage. Our results provide new theoretical justifications for (Bayesian) deep learning with ReLU activation functions, highlighting their inferential potential.
Published 2020-02-26
URL https://arxiv.org/abs/2002.11815v1
PDF https://arxiv.org/pdf/2002.11815v1.pdf
PWC https://paperswithcode.com/paper/uncertainty-quantification-for-sparse-deep

An Online Learning Framework for Energy-Efficient Navigation of Electric Vehicles

Title An Online Learning Framework for Energy-Efficient Navigation of Electric Vehicles
Authors Niklas Åkerblom, Yuxin Chen, Morteza Haghir Chehreghani
Abstract Energy-efficient navigation constitutes an important challenge in electric vehicles, due to their limited battery capacity. We employ a Bayesian approach to model energy consumption at road-segments for efficient navigation. In order to learn the model parameters, we develop an online learning framework and investigate several exploration strategies such as Thompson Sampling and Upper Confidence Bound. We then extend our online learning framework to multi-agent setting, where multiple vehicles adaptively navigate and learn the parameters of the energy model. We analyze Thompson Sampling and establish rigorous regret bounds on its performance. Finally, we demonstrate the performance of our methods via several real-world experiments on Luxembourg SUMO Traffic dataset.
Published 2020-03-03
URL https://arxiv.org/abs/2003.01416v1
PDF https://arxiv.org/pdf/2003.01416v1.pdf
PWC https://paperswithcode.com/paper/an-online-learning-framework-for-energy

Choice functions based on sets of strict partial orders: an axiomatic characterisation

Title Choice functions based on sets of strict partial orders: an axiomatic characterisation
Authors Jasper De Bock
Abstract Methods for choosing from a set of options are often based on a strict partial order on these options, or on a set of such partial orders. I here provide a very general axiomatic characterisation for choice functions of this form. It includes as special cases an axiomatic characterisation for choice functions based on (sets of) total orders, (sets of) weak orders, (sets of) coherent lower previsions and (sets of) probability measures.
Published 2020-03-25
URL https://arxiv.org/abs/2003.11631v1
PDF https://arxiv.org/pdf/2003.11631v1.pdf
PWC https://paperswithcode.com/paper/choice-functions-based-on-sets-of-strict

Text Extraction and Restoration of Old Handwritten Documents

Title Text Extraction and Restoration of Old Handwritten Documents
Authors Mayank Wadhwani, Debapriya Kundu, Deepayan Chakraborty, Bhabatosh Chanda
Abstract Image restoration is very crucial computer vision task. This paper describes two novel methods for the restoration of old degraded handwritten documents using deep neural network. In addition to that, a small-scale dataset of 26 heritage letters images is introduced. The ground truth data to train the desired network is generated semi automatically involving a pragmatic combination of color transformation, Gaussian mixture model based segmentation and shape correction by using mathematical morphological operators. In the first approach, a deep neural network has been used for text extraction from the document image and later background reconstruction has been done using Gaussian mixture modeling. But Gaussian mixture modelling requires to set parameters manually, to alleviate this we propose a second approach where the background reconstruction and foreground extraction (which which includes extracting text with its original colour) both has been done using deep neural network. Experiments demonstrate that the proposed systems perform well on handwritten document images with severe degradations, even when trained with small dataset. Hence, the proposed methods are ideally suited for digital heritage preservation repositories. It is worth mentioning that, these methods can be extended easily for printed degraded documents.
Tasks Image Restoration
Published 2020-01-23
URL https://arxiv.org/abs/2001.08742v1
PDF https://arxiv.org/pdf/2001.08742v1.pdf
PWC https://paperswithcode.com/paper/text-extraction-and-restoration-of-old

Learning CHARME models with (deep) neural networks

Title Learning CHARME models with (deep) neural networks
Authors José G. Gómez-García, Jalal Fadili, Christophe Chesneau
Abstract In this paper, we consider a model called CHARME (Conditional Heteroscedastic Autoregressive Mixture of Experts), a class of generalized mixture of nonlinear nonparametric AR-ARCH time series. Under certain Lipschitz-type conditions on the autoregressive and volatility functions, we prove that this model is stationary, ergodic and $\tau$-weakly dependent. These conditions are much weaker than those presented in the literature that treats this model. Moreover, this result forms the theoretical basis for deriving an asymptotic theory of the underlying (non)parametric estimation, which we present for this model. As an application, from the universal approximation property of neural networks (NN), possibly with deep architectures, we develop a learning theory for the NN-based autoregressive functions of the model, where the strong consistency and asymptotic normality of the considered estimator of the NN weights and biases are guaranteed under weak conditions.
Tasks Time Series
Published 2020-02-08
URL https://arxiv.org/abs/2002.03237v1
PDF https://arxiv.org/pdf/2002.03237v1.pdf
PWC https://paperswithcode.com/paper/learning-charme-models-with-deep-neural

Uncovering differential equations from data with hidden variables

Title Uncovering differential equations from data with hidden variables
Authors Agustín Somacal, Leonardo Boechi, Matthieu Jonckheere, Vincent Lefieux, Dominique Picard, Ezequiel Smucler
Abstract Finding a set of differential equations to model dynamical systems is a difficult task present in many branches of science and engineering. We propose a method to learn systems of differential equations directly from data. Our method is based on solving a tailor-made $\ell_1-$regularised least-squares problem and can deal with latent variables by adding higher-order derivatives to account for the lack of information. Extensive numerical studies show that our method can recover useful representations of the dynamical system that generated the data even when some variables are not observed. Moreover, being based on solving a convex optimisation problem, our method is much faster than competing approaches based on solving combinatorial problems. Finally, we apply our methodology to a real data-set of temperature time series.
Tasks Time Series
Published 2020-02-06
URL https://arxiv.org/abs/2002.02250v1
PDF https://arxiv.org/pdf/2002.02250v1.pdf
PWC https://paperswithcode.com/paper/uncovering-differential-equations-from-data

Linear time dynamic programming for the exact path of optimal models selected from a finite set

Title Linear time dynamic programming for the exact path of optimal models selected from a finite set
Authors Toby Hocking, Joseph Vargovich
Abstract Many learning algorithms are formulated in terms of finding model parameters which minimize a data-fitting loss function plus a regularizer. When the regularizer involves the l0 pseudo-norm, the resulting regularization path consists of a finite set of models. The fastest existing algorithm for computing the breakpoints in the regularization path is quadratic in the number of models, so it scales poorly to high dimensional problems. We provide new formal proofs that a dynamic programming algorithm can be used to compute the breakpoints in linear time. Empirical results on changepoint detection problems demonstrate the improved accuracy and speed relative to grid search and the previous quadratic time algorithm.
Published 2020-03-05
URL https://arxiv.org/abs/2003.02808v1
PDF https://arxiv.org/pdf/2003.02808v1.pdf
PWC https://paperswithcode.com/paper/linear-time-dynamic-programming-for-the-exact
comments powered by Disqus