July 29, 2019

2848 words 14 mins read

Paper Group ANR 6

Deep GrabCut for Object Selection. Superposition de calques monochromes d’opacités variables. End-to-End Attention based Text-Dependent Speaker Verification. Unsupervised Learning via Total Correlation Explanation. ProLanGO: Protein Function Prediction Using Neural~Machine Translation Based on a Recurrent Neural Network. Semantic Refinement GRU-bas …

Deep GrabCut for Object Selection


Title	Deep GrabCut for Object Selection
Authors	Ning Xu, Brian Price, Scott Cohen, Jimei Yang, Thomas Huang
Abstract	Most previous bounding-box-based segmentation methods assume the bounding box tightly covers the object of interest. However it is common that a rectangle input could be too large or too small. In this paper, we propose a novel segmentation approach that uses a rectangle as a soft constraint by transforming it into an Euclidean distance map. A convolutional encoder-decoder network is trained end-to-end by concatenating images with these distance maps as inputs and predicting the object masks as outputs. Our approach gets accurate segmentation results given sloppy rectangles while being general for both interactive segmentation and instance segmentation. We show our network extends to curve-based input without retraining. We further apply our network to instance-level semantic segmentation and resolve any overlap using a conditional random field. Experiments on benchmark datasets demonstrate the effectiveness of the proposed approaches.
Tasks	Instance Segmentation, Interactive Segmentation, Semantic Segmentation
Published	2017-07-02
URL	http://arxiv.org/abs/1707.00243v2
PDF	http://arxiv.org/pdf/1707.00243v2.pdf
PWC	https://paperswithcode.com/paper/deep-grabcut-for-object-selection
Repo
Framework

Superposition de calques monochromes d’opacités variables


Title	Superposition de calques monochromes d’opacités variables
Authors	Alexandre Bali
Abstract	For a monochrome layer $x$ of opacity $0\le o_x\le1 $ placed on another monochrome layer of opacity 1, the result given by the standard formula is $$\small\Pi\left({\bf C}\varphi\right)=1+\sum{n=1}^2\left(2-n-(-1)^no_{\chi(\varphi+1)}\right)\left(\chi(n+\varphi-1)-o_{\chi(n+\varphi-1)}\right),$$ the formula being of course explained in detail in this paper. We will eventually deduce a very simple theorem, generalize it and then see its validity with alternative formulas to this standard containing the same main properties here exposed.
Tasks
Published	2017-07-13
URL	https://arxiv.org/abs/1707.09839v3
PDF	https://arxiv.org/pdf/1707.09839v3.pdf
PWC	https://paperswithcode.com/paper/superposition-de-calques-monochromes
Repo
Framework

End-to-End Attention based Text-Dependent Speaker Verification


Title	End-to-End Attention based Text-Dependent Speaker Verification
Authors	Shi-Xiong Zhang, Zhuo Chen, Yong Zhao, Jinyu Li, Yifan Gong
Abstract	A new type of End-to-End system for text-dependent speaker verification is presented in this paper. Previously, using the phonetically discriminative/speaker discriminative DNNs as feature extractors for speaker verification has shown promising results. The extracted frame-level (DNN bottleneck, posterior or d-vector) features are equally weighted and aggregated to compute an utterance-level speaker representation (d-vector or i-vector). In this work we use speaker discriminative CNNs to extract the noise-robust frame-level features. These features are smartly combined to form an utterance-level speaker vector through an attention mechanism. The proposed attention model takes the speaker discriminative information and the phonetic information to learn the weights. The whole system, including the CNN and attention model, is joint optimized using an end-to-end criterion. The training algorithm imitates exactly the evaluation process — directly mapping a test utterance and a few target speaker utterances into a single verification score. The algorithm can automatically select the most similar impostor for each target speaker to train the network. We demonstrated the effectiveness of the proposed end-to-end system on Windows $10$ “Hey Cortana” speaker verification task.
Tasks	Speaker Verification, Text-Dependent Speaker Verification
Published	2017-01-03
URL	http://arxiv.org/abs/1701.00562v1
PDF	http://arxiv.org/pdf/1701.00562v1.pdf
PWC	https://paperswithcode.com/paper/end-to-end-attention-based-text-dependent
Repo
Framework

Unsupervised Learning via Total Correlation Explanation


Title	Unsupervised Learning via Total Correlation Explanation
Authors	Greg Ver Steeg
Abstract	Learning by children and animals occurs effortlessly and largely without obvious supervision. Successes in automating supervised learning have not translated to the more ambiguous realm of unsupervised learning where goals and labels are not provided. Barlow (1961) suggested that the signal that brains leverage for unsupervised learning is dependence, or redundancy, in the sensory environment. Dependence can be characterized using the information-theoretic multivariate mutual information measure called total correlation. The principle of Total Cor-relation Ex-planation (CorEx) is to learn representations of data that “explain” as much dependence in the data as possible. We review some manifestations of this principle along with successes in unsupervised learning problems across diverse domains including human behavior, biology, and language.
Tasks
Published	2017-06-27
URL	http://arxiv.org/abs/1706.08984v1
PDF	http://arxiv.org/pdf/1706.08984v1.pdf
PWC	https://paperswithcode.com/paper/unsupervised-learning-via-total-correlation
Repo
Framework

ProLanGO: Protein Function Prediction Using Neural~Machine Translation Based on a Recurrent Neural Network


Title	ProLanGO: Protein Function Prediction Using Neural~Machine Translation Based on a Recurrent Neural Network
Authors	Renzhi Cao, Colton Freitas, Leong Chan, Miao Sun, Haiqing Jiang, Zhangxin Chen
Abstract	With the development of next generation sequencing techniques, it is fast and cheap to determine protein sequences but relatively slow and expensive to extract useful information from protein sequences because of limitations of traditional biological experimental techniques. Protein function prediction has been a long standing challenge to fill the gap between the huge amount of protein sequences and the known function. In this paper, we propose a novel method to convert the protein function problem into a language translation problem by the new proposed protein sequence language “ProLan” to the protein function language “GOLan”, and build a neural machine translation model based on recurrent neural networks to translate “ProLan” language to “GOLan” language. We blindly tested our method by attending the latest third Critical Assessment of Function Annotation (CAFA 3) in 2016, and also evaluate the performance of our methods on selected proteins whose function was released after CAFA competition. The good performance on the training and testing datasets demonstrates that our new proposed method is a promising direction for protein function prediction. In summary, we first time propose a method which converts the protein function prediction problem to a language translation problem and applies a neural machine translation model for protein function prediction.
Tasks	Machine Translation, Protein Function Prediction
Published	2017-10-19
URL	http://arxiv.org/abs/1710.07016v1
PDF	http://arxiv.org/pdf/1710.07016v1.pdf
PWC	https://paperswithcode.com/paper/prolango-protein-function-prediction-using
Repo
Framework


Title	Semantic Refinement GRU-based Neural Language Generation for Spoken Dialogue Systems
Authors	Van-Khanh Tran, Le-Minh Nguyen
Abstract	Natural language generation (NLG) plays a critical role in spoken dialogue systems. This paper presents a new approach to NLG by using recurrent neural networks (RNN), in which a gating mechanism is applied before RNN computation. This allows the proposed model to generate appropriate sentences. The RNN-based generator can be learned from unaligned data by jointly training sentence planning and surface realization to produce natural language responses. The model was extensively evaluated on four different NLG domains. The results show that the proposed generator achieved better performance on all the NLG domains compared to previous generators.
Tasks	Spoken Dialogue Systems, Text Generation
Published	2017-06-01
URL	http://arxiv.org/abs/1706.00134v4
PDF	http://arxiv.org/pdf/1706.00134v4.pdf
PWC	https://paperswithcode.com/paper/semantic-refinement-gru-based-neural-language
Repo
Framework

Identification of Gaussian Process State Space Models


Title	Identification of Gaussian Process State Space Models
Authors	Stefanos Eleftheriadis, Thomas F. W. Nicholson, Marc Peter Deisenroth, James Hensman
Abstract	The Gaussian process state space model (GPSSM) is a non-linear dynamical system, where unknown transition and/or measurement mappings are described by GPs. Most research in GPSSMs has focussed on the state estimation problem, i.e., computing a posterior of the latent state given the model. However, the key challenge in GPSSMs has not been satisfactorily addressed yet: system identification, i.e., learning the model. To address this challenge, we impose a structured Gaussian variational posterior distribution over the latent states, which is parameterised by a recognition model in the form of a bi-directional recurrent neural network. Inference with this structure allows us to recover a posterior smoothed over sequences of data. We provide a practical algorithm for efficiently computing a lower bound on the marginal likelihood using the reparameterisation trick. This further allows for the use of arbitrary kernels within the GPSSM. We demonstrate that the learnt GPSSM can efficiently generate plausible future trajectories of the identified system after only observing a small number of episodes from the true system.
Tasks
Published	2017-05-30
URL	http://arxiv.org/abs/1705.10888v2
PDF	http://arxiv.org/pdf/1705.10888v2.pdf
PWC	https://paperswithcode.com/paper/identification-of-gaussian-process-state-1
Repo
Framework

Deep Recurrent Neural Network for Protein Function Prediction from Sequence


Title	Deep Recurrent Neural Network for Protein Function Prediction from Sequence
Authors	Xueliang Liu
Abstract	As high-throughput biological sequencing becomes faster and cheaper, the need to extract useful information from sequencing becomes ever more paramount, often limited by low-throughput experimental characterizations. For proteins, accurate prediction of their functions directly from their primary amino-acid sequences has been a long standing challenge. Here, machine learning using artificial recurrent neural networks (RNN) was applied towards classification of protein function directly from primary sequence without sequence alignment, heuristic scoring or feature engineering. The RNN models containing long-short-term-memory (LSTM) units trained on public, annotated datasets from UniProt achieved high performance for in-class prediction of four important protein functions tested, particularly compared to other machine learning algorithms using sequence-derived protein features. RNN models were used also for out-of-class predictions of phylogenetically distinct protein families with similar functions, including proteins of the CRISPR-associated nuclease, ferritin-like iron storage and cytochrome P450 families. Applying the trained RNN models on the partially unannotated UniRef100 database predicted not only candidates validated by existing annotations but also currently unannotated sequences. Some RNN predictions for the ferritin-like iron sequestering function were experimentally validated, even though their sequences differ significantly from known, characterized proteins and from each other and cannot be easily predicted using popular bioinformatics methods. As sequencing and experimental characterization data increases rapidly, the machine-learning approach based on RNN could be useful for discovery and prediction of homologues for a wide range of protein functions.
Tasks	Feature Engineering, Protein Function Prediction
Published	2017-01-28
URL	http://arxiv.org/abs/1701.08318v1
PDF	http://arxiv.org/pdf/1701.08318v1.pdf
PWC	https://paperswithcode.com/paper/deep-recurrent-neural-network-for-protein
Repo
Framework

Default Logic and Bounded Treewidth


Title	Default Logic and Bounded Treewidth
Authors	Johannes K. Fichte, Markus Hecher, Irina Schindler
Abstract	In this paper, we study Reiter’s propositional default logic when the treewidth of a certain graph representation (semi-primal graph) of the input theory is bounded. We establish a dynamic programming algorithm on tree decompositions that decides whether a theory has a consistent stable extension (Ext). Our algorithm can even be used to enumerate all generating defaults (ExtEnum) that lead to stable extensions. We show that our algorithm decides Ext in linear time in the input theory and triple exponential time in the treewidth (so-called fixed-parameter linear algorithm). Further, our algorithm solves ExtEnum with a pre-computation step that is linear in the input theory and triple exponential in the treewidth followed by a linear delay to output solutions.
Tasks
Published	2017-06-28
URL	http://arxiv.org/abs/1706.09393v2
PDF	http://arxiv.org/pdf/1706.09393v2.pdf
PWC	https://paperswithcode.com/paper/default-logic-and-bounded-treewidth
Repo
Framework

A Connection between Feed-Forward Neural Networks and Probabilistic Graphical Models


Title	A Connection between Feed-Forward Neural Networks and Probabilistic Graphical Models
Authors	Dmitrij Schlesinger
Abstract	Two of the most popular modelling paradigms in computer vision are feed-forward neural networks (FFNs) and probabilistic graphical models (GMs). Various connections between the two have been studied in recent works, such as e.g. expressing mean-field based inference in a GM as an FFN. This paper establishes a new connection between FFNs and GMs. Our key observation is that any FFN implements a certain approximation of a corresponding Bayesian network (BN). We characterize various benefits of having this connection. In particular, it results in a new learning algorithm for BNs. We validate the proposed methods for a classification problem on CIFAR-10 dataset and for binary image segmentation on Weizmann Horse dataset. We show that statistically learned BNs improve performance, having at the same time essentially better generalization capability, than their FFN counterparts.
Tasks	Semantic Segmentation
Published	2017-10-30
URL	http://arxiv.org/abs/1710.11052v1
PDF	http://arxiv.org/pdf/1710.11052v1.pdf
PWC	https://paperswithcode.com/paper/a-connection-between-feed-forward-neural
Repo
Framework

Adversarial Examples for Evaluating Reading Comprehension Systems


Title	Adversarial Examples for Evaluating Reading Comprehension Systems
Authors	Robin Jia, Percy Liang
Abstract	Standard accuracy metrics indicate that reading comprehension systems are making rapid progress, but the extent to which these systems truly understand language remains unclear. To reward systems with real language understanding abilities, we propose an adversarial evaluation scheme for the Stanford Question Answering Dataset (SQuAD). Our method tests whether systems can answer questions about paragraphs that contain adversarially inserted sentences, which are automatically generated to distract computer systems without changing the correct answer or misleading humans. In this adversarial setting, the accuracy of sixteen published models drops from an average of $75%$ F1 score to $36%$; when the adversary is allowed to add ungrammatical sequences of words, average accuracy on four models decreases further to $7%$. We hope our insights will motivate the development of new models that understand language more precisely.
Tasks	Accuracy Metrics, Question Answering, Reading Comprehension
Published	2017-07-23
URL	http://arxiv.org/abs/1707.07328v1
PDF	http://arxiv.org/pdf/1707.07328v1.pdf
PWC	https://paperswithcode.com/paper/adversarial-examples-for-evaluating-reading
Repo
Framework

Towards learning domain-independent planning heuristics


Title	Towards learning domain-independent planning heuristics
Authors	Pawel Gomoluch, Dalal Alrajeh, Alessandra Russo, Antonio Bucchiarone
Abstract	Automated planning remains one of the most general paradigms in Artificial Intelligence, providing means of solving problems coming from a wide variety of domains. One of the key factors restricting the applicability of planning is its computational complexity resulting from exponentially large search spaces. Heuristic approaches are necessary to solve all but the simplest problems. In this work, we explore the possibility of obtaining domain-independent heuristic functions using machine learning. This is a part of a wider research program whose objective is to improve practical applicability of planning in systems for which the planning domains evolve at run time. The challenge is therefore the learning of (corrections of) domain-independent heuristics that can be reused across different planning domains.
Tasks
Published	2017-07-21
URL	http://arxiv.org/abs/1707.06895v1
PDF	http://arxiv.org/pdf/1707.06895v1.pdf
PWC	https://paperswithcode.com/paper/towards-learning-domain-independent-planning
Repo
Framework

Tree Ensembles with Rule Structured Horseshoe Regularization


Title	Tree Ensembles with Rule Structured Horseshoe Regularization
Authors	Malte Nalenz, Mattias Villani
Abstract	We propose a new Bayesian model for flexible nonlinear regression and classification using tree ensembles. The model is based on the RuleFit approach in Friedman and Popescu (2008) where rules from decision trees and linear terms are used in a L1-regularized regression. We modify RuleFit by replacing the L1-regularization by a horseshoe prior, which is well known to give aggressive shrinkage of noise predictor while leaving the important signal essentially untouched. This is especially important when a large number of rules are used as predictors as many of them only contribute noise. Our horseshoe prior has an additional hierarchical layer that applies more shrinkage a priori to rules with a large number of splits, and to rules that are only satisfied by a few observations. The aggressive noise shrinkage of our prior also makes it possible to complement the rules from boosting in Friedman and Popescu (2008) with an additional set of trees from random forest, which brings a desirable diversity to the ensemble. We sample from the posterior distribution using a very efficient and easily implemented Gibbs sampler. The new model is shown to outperform state-of-the-art methods like RuleFit, BART and random forest on 16 datasets. The model and its interpretation is demonstrated on the well known Boston housing data, and on gene expression data for cancer classification. The posterior sampling, prediction and graphical tools for interpreting the model results are implemented in a publicly available R package.
Tasks
Published	2017-02-16
URL	http://arxiv.org/abs/1702.05008v2
PDF	http://arxiv.org/pdf/1702.05008v2.pdf
PWC	https://paperswithcode.com/paper/tree-ensembles-with-rule-structured-horseshoe
Repo
Framework

Self-Supervised Intrinsic Image Decomposition


Title	Self-Supervised Intrinsic Image Decomposition
Authors	Michael Janner, Jiajun Wu, Tejas D. Kulkarni, Ilker Yildirim, Joshua B. Tenenbaum
Abstract	Intrinsic decomposition from a single image is a highly challenging task, due to its inherent ambiguity and the scarcity of training data. In contrast to traditional fully supervised learning approaches, in this paper we propose learning intrinsic image decomposition by explaining the input image. Our model, the Rendered Intrinsics Network (RIN), joins together an image decomposition pipeline, which predicts reflectance, shape, and lighting conditions given a single image, with a recombination function, a learned shading model used to recompose the original input based off of intrinsic image predictions. Our network can then use unsupervised reconstruction error as an additional signal to improve its intermediate representations. This allows large-scale unlabeled data to be useful during training, and also enables transferring learned knowledge to images of unseen object categories, lighting conditions, and shapes. Extensive experiments demonstrate that our method performs well on both intrinsic image decomposition and knowledge transfer.
Tasks	Intrinsic Image Decomposition, Transfer Learning
Published	2017-11-10
URL	http://arxiv.org/abs/1711.03678v2
PDF	http://arxiv.org/pdf/1711.03678v2.pdf
PWC	https://paperswithcode.com/paper/self-supervised-intrinsic-image-decomposition
Repo
Framework

A novel image tag completion method based on convolutional neural network


Title	A novel image tag completion method based on convolutional neural network
Authors	Yanyan Geng, Guohui Zhang, Weizhi Li, Yi Gu, Ru-Ze Liang, Gaoyuan Liang, Jingbin Wang, Yanbin Wu, Nitin Patil, Jing-Yan Wang
Abstract	In the problems of image retrieval and annotation, complete textual tag lists of images play critical roles. However, in real-world applications, the image tags are usually incomplete, thus it is important to learn the complete tags for images. In this paper, we study the problem of image tag complete and proposed a novel method for this problem based on a popular image representation method, convolutional neural network (CNN). The method estimates the complete tags from the convolutional filtering outputs of images based on a linear predictor. The CNN parameters, linear predictor, and the complete tags are learned jointly by our method. We build a minimization problem to encourage the consistency between the complete tags and the available incomplete tags, reduce the estimation error, and reduce the model complexity. An iterative algorithm is developed to solve the minimization problem. Experiments over benchmark image data sets show its effectiveness.
Tasks	Image Retrieval
Published	2017-03-02
URL	http://arxiv.org/abs/1703.00586v2
PDF	http://arxiv.org/pdf/1703.00586v2.pdf
PWC	https://paperswithcode.com/paper/a-novel-image-tag-completion-method-based-on
Repo
Framework