July 27, 2019

3453 words 17 mins read

Paper Group ANR 747

The Heterogeneous Ensembles of Standard Classification Algorithms (HESCA): the Whole is Greater than the Sum of its Parts. 3D Object Discovery and Modeling Using Single RGB-D Images Containing Multiple Object Instances. Stage 4 validation of the Satellite Image Automatic Mapper lightweight computer program for Earth observation Level 2 product gene …

The Heterogeneous Ensembles of Standard Classification Algorithms (HESCA): the Whole is Greater than the Sum of its Parts


Title	The Heterogeneous Ensembles of Standard Classification Algorithms (HESCA): the Whole is Greater than the Sum of its Parts
Authors	James Large, Jason Lines, Anthony Bagnall
Abstract	Building classification models is an intrinsically practical exercise that requires many design decisions prior to deployment. We aim to provide some guidance in this decision making process. Specifically, given a classification problem with real valued attributes, we consider which classifier or family of classifiers should one use. Strong contenders are tree based homogeneous ensembles, support vector machines or deep neural networks. All three families of model could claim to be state-of-the-art, and yet it is not clear when one is preferable to the others. Our extensive experiments with over 200 data sets from two distinct archives demonstrate that, rather than choose a single family and expend computing resources on optimising that model, it is significantly better to build simpler versions of classifiers from each family and ensemble. We show that the Heterogeneous Ensembles of Standard Classification Algorithms (HESCA), which ensembles based on error estimates formed on the train data, is significantly better (in terms of error, balanced error, negative log likelihood and area under the ROC curve) than its individual components, picking the component that is best on train data, and a support vector machine tuned over 1089 different parameter configurations. We demonstrate HESCA+, which contains a deep neural network, a support vector machine and two decision tree forests, is significantly better than its components, picking the best component, and HESCA. We analyse the results further and find that HESCA and HESCA+ are of particular value when the train set size is relatively small and the problem has multiple classes. HESCA is a fast approach that is, on average, as good as state-of-the-art classifiers, whereas HESCA+ is significantly better than average and represents a strong benchmark for future research.
Tasks	Decision Making
Published	2017-10-25
URL	http://arxiv.org/abs/1710.09220v1
PDF	http://arxiv.org/pdf/1710.09220v1.pdf
PWC	https://paperswithcode.com/paper/the-heterogeneous-ensembles-of-standard
Repo
Framework

3D Object Discovery and Modeling Using Single RGB-D Images Containing Multiple Object Instances


Title	3D Object Discovery and Modeling Using Single RGB-D Images Containing Multiple Object Instances
Authors	Wim Abbeloos, Esra Ataer-Cansizoglu, Sergio Caccamo, Yuichi Taguchi, Yukiyasu Domae
Abstract	Unsupervised object modeling is important in robotics, especially for handling a large set of objects. We present a method for unsupervised 3D object discovery, reconstruction, and localization that exploits multiple instances of an identical object contained in a single RGB-D image. The proposed method does not rely on segmentation, scene knowledge, or user input, and thus is easily scalable. Our method aims to find recurrent patterns in a single RGB-D image by utilizing appearance and geometry of the salient regions. We extract keypoints and match them in pairs based on their descriptors. We then generate triplets of the keypoints matching with each other using several geometric criteria to minimize false matches. The relative poses of the matched triplets are computed and clustered to discover sets of triplet pairs with similar relative poses. Triplets belonging to the same set are likely to belong to the same object and are used to construct an initial object model. Detection of remaining instances with the initial object model using RANSAC allows to further expand and refine the model. The automatically generated object models are both compact and descriptive. We show quantitative and qualitative results on RGB-D images with various objects including some from the Amazon Picking Challenge. We also demonstrate the use of our method in an object picking scenario with a robotic arm.
Tasks
Published	2017-10-17
URL	http://arxiv.org/abs/1710.06231v1
PDF	http://arxiv.org/pdf/1710.06231v1.pdf
PWC	https://paperswithcode.com/paper/3d-object-discovery-and-modeling-using-single
Repo
Framework

Stage 4 validation of the Satellite Image Automatic Mapper lightweight computer program for Earth observation Level 2 product generation, Part 1 Theory


Title	Stage 4 validation of the Satellite Image Automatic Mapper lightweight computer program for Earth observation Level 2 product generation, Part 1 Theory
Authors	Andrea Baraldi, Michael Laurence Humber, Dirk Tiede, Stefan Lang
Abstract	The European Space Agency (ESA) defines an Earth Observation (EO) Level 2 product as a multispectral (MS) image corrected for geometric, atmospheric, adjacency and topographic effects, stacked with its scene classification map (SCM), whose legend includes quality layers such as cloud and cloud-shadow. No ESA EO Level 2 product has ever been systematically generated at the ground segment. To contribute toward filling an information gap from EO big data to the ESA EO Level 2 product, an original Stage 4 validation (Val) of the Satellite Image Automatic Mapper (SIAM) lightweight computer program was conducted by independent means on an annual Web-Enabled Landsat Data (WELD) image composite time-series of the conterminous U.S. The core of SIAM is a one pass prior knowledge based decision tree for MS reflectance space hyperpolyhedralization into static color names presented in literature in recent years. For the sake of readability this paper is split into two. The present Part 1 Theory provides the multidisciplinary background of a priori color naming in cognitive science, from linguistics to computer vision. To cope with dictionaries of MS color names and land cover class names that do not coincide and must be harmonized, an original hybrid guideline is proposed to identify a categorical variable pair relationship. An original quantitative measure of categorical variable pair association is also proposed. The subsequent Part 2 Validation discusses Stage 4 Val results collected by an original protocol for wall-to-wall thematic map quality assessment without sampling where the test and reference map legends can differ. Conclusions are that the SIAM-WELD maps instantiate a Level 2 SCM product whose legend is the 4 class taxonomy of the FAO Land Cover Classification System at the Dichotomous Phase Level 1 vegetation/nonvegetation and Level 2 terrestrial/aquatic.
Tasks	Scene Classification, Time Series
Published	2017-01-08
URL	http://arxiv.org/abs/1701.01930v1
PDF	http://arxiv.org/pdf/1701.01930v1.pdf
PWC	https://paperswithcode.com/paper/stage-4-validation-of-the-satellite-image
Repo
Framework

Learning to segment with image-level supervision


Title	Learning to segment with image-level supervision
Authors	Gaurav Pandey, Ambedkar Dukkipati
Abstract	Deep convolutional networks have achieved the state-of-the-art for semantic image segmentation tasks. However, training these networks requires access to densely labeled images, which are known to be very expensive to obtain. On the other hand, the web provides an almost unlimited source of images annotated at the image level. How can one utilize this much larger weakly annotated set for tasks that require dense labeling? Prior work often relied on localization cues, such as saliency maps, objectness priors, bounding boxes etc., to address this challenging problem. In this paper, we propose a model that generates auxiliary labels for each image, while simultaneously forcing the output of the CNN to satisfy the mean-field constraints imposed by a conditional random field. We show that one can enforce the CRF constraints by forcing the distribution at each pixel to be close to the distribution of its neighbors. This is in stark contrast with methods that compute a recursive expansion of the mean-field distribution using a recurrent architecture and train the resultant distribution. Instead, the proposed model adds an extra loss term to the output of the CNN, and hence, is faster than recursive implementations. We achieve the state-of-the-art for weakly supervised semantic image segmentation on VOC 2012 dataset, assuming no manually labeled pixel level information is available. Furthermore, the incorporation of conditional random fields in CNN incurs little extra time during training.
Tasks	Semantic Segmentation
Published	2017-05-03
URL	http://arxiv.org/abs/1705.01262v2
PDF	http://arxiv.org/pdf/1705.01262v2.pdf
PWC	https://paperswithcode.com/paper/learning-to-segment-with-image-level
Repo
Framework

Deep Long Short-Term Memory Adaptive Beamforming Networks For Multichannel Robust Speech Recognition


Title	Deep Long Short-Term Memory Adaptive Beamforming Networks For Multichannel Robust Speech Recognition
Authors	Zhong Meng, Shinji Watanabe, John R. Hershey, Hakan Erdogan
Abstract	Far-field speech recognition in noisy and reverberant conditions remains a challenging problem despite recent deep learning breakthroughs. This problem is commonly addressed by acquiring a speech signal from multiple microphones and performing beamforming over them. In this paper, we propose to use a recurrent neural network with long short-term memory (LSTM) architecture to adaptively estimate real-time beamforming filter coefficients to cope with non-stationary environmental noise and dynamic nature of source and microphones positions which results in a set of timevarying room impulse responses. The LSTM adaptive beamformer is jointly trained with a deep LSTM acoustic model to predict senone labels. Further, we use hidden units in the deep LSTM acoustic model to assist in predicting the beamforming filter coefficients. The proposed system achieves 7.97% absolute gain over baseline systems with no beamforming on CHiME-3 real evaluation set.
Tasks	Robust Speech Recognition, Speech Recognition
Published	2017-11-21
URL	http://arxiv.org/abs/1711.08016v1
PDF	http://arxiv.org/pdf/1711.08016v1.pdf
PWC	https://paperswithcode.com/paper/deep-long-short-term-memory-adaptive
Repo
Framework

Learning causal effects from many randomized experiments using regularized instrumental variables


Title	Learning causal effects from many randomized experiments using regularized instrumental variables
Authors	Alexander Peysakhovich, Dean Eckles
Abstract	Scientific and business practices are increasingly resulting in large collections of randomized experiments. Analyzed together, these collections can tell us things that individual experiments in the collection cannot. We study how to learn causal relationships between variables from the kinds of collections faced by modern data scientists: the number of experiments is large, many experiments have very small effects, and the analyst lacks metadata (e.g., descriptions of the interventions). Here we use experimental groups as instrumental variables (IV) and show that a standard method (two-stage least squares) is biased even when the number of experiments is infinite. We show how a sparsity-inducing l0 regularization can — in a reversal of the standard bias–variance tradeoff in regularization — reduce bias (and thus error) of interventional predictions. Because we are interested in interventional loss minimization we also propose a modified cross-validation procedure (IVCV) to feasibly select the regularization parameter. We show, using a trick from Monte Carlo sampling, that IVCV can be done using summary statistics instead of raw data. This makes our full procedure simple to use in many real-world applications.
Tasks
Published	2017-01-04
URL	http://arxiv.org/abs/1701.01140v3
PDF	http://arxiv.org/pdf/1701.01140v3.pdf
PWC	https://paperswithcode.com/paper/learning-causal-effects-from-many-randomized
Repo
Framework

Two forms of minimality in ASPIC+


Title	Two forms of minimality in ASPIC+
Authors	Zimi Li, Andrea Cohen, Simon Parsons
Abstract	Many systems of structured argumentation explicitly require that the facts and rules that make up the argument for a conclusion be the minimal set required to derive the conclusion. ASPIC+ does not place such a requirement on arguments, instead requiring that every rule and fact that are part of an argument be used in its construction. Thus ASPIC+ arguments are minimal in the sense that removing any element of the argument would lead to a structure that is not an argument. In this brief note we discuss these two types of minimality and show how the first kind of minimality can, if desired, be recovered in ASPIC+.
Tasks
Published	2017-02-02
URL	http://arxiv.org/abs/1702.00780v1
PDF	http://arxiv.org/pdf/1702.00780v1.pdf
PWC	https://paperswithcode.com/paper/two-forms-of-minimality-in-aspic
Repo
Framework

Unsupervised Domain Adaptation for Face Recognition in Unlabeled Videos


Title	Unsupervised Domain Adaptation for Face Recognition in Unlabeled Videos
Authors	Kihyuk Sohn, Sifei Liu, Guangyu Zhong, Xiang Yu, Ming-Hsuan Yang, Manmohan Chandraker
Abstract	Despite rapid advances in face recognition, there remains a clear gap between the performance of still image-based face recognition and video-based face recognition, due to the vast difference in visual quality between the domains and the difficulty of curating diverse large-scale video datasets. This paper addresses both of those challenges, through an image to video feature-level domain adaptation approach, to learn discriminative video frame representations. The framework utilizes large-scale unlabeled video data to reduce the gap between different domains while transferring discriminative knowledge from large-scale labeled still images. Given a face recognition network that is pretrained in the image domain, the adaptation is achieved by (i) distilling knowledge from the network to a video adaptation network through feature matching, (ii) performing feature restoration through synthetic data augmentation and (iii) learning a domain-invariant feature through a domain adversarial discriminator. We further improve performance through a discriminator-guided feature fusion that boosts high-quality frames while eliminating those degraded by video domain-specific factors. Experiments on the YouTube Faces and IJB-A datasets demonstrate that each module contributes to our feature-level domain adaptation framework and substantially improves video face recognition performance to achieve state-of-the-art accuracy. We demonstrate qualitatively that the network learns to suppress diverse artifacts in videos such as pose, illumination or occlusion without being explicitly trained for them.
Tasks	Data Augmentation, Domain Adaptation, Face Recognition, Unsupervised Domain Adaptation
Published	2017-08-07
URL	http://arxiv.org/abs/1708.02191v1
PDF	http://arxiv.org/pdf/1708.02191v1.pdf
PWC	https://paperswithcode.com/paper/unsupervised-domain-adaptation-for-face
Repo
Framework

Learning Less-Overlapping Representations


Title	Learning Less-Overlapping Representations
Authors	Pengtao Xie, Hongbao Zhang, Eric P. Xing
Abstract	In representation learning (RL), how to make the learned representations easy to interpret and less overfitted to training data are two important but challenging issues. To address these problems, we study a new type of regulariza- tion approach that encourages the supports of weight vectors in RL models to have small overlap, by simultaneously promoting near-orthogonality among vectors and sparsity of each vector. We apply the proposed regularizer to two models: neural networks (NNs) and sparse coding (SC), and develop an efficient ADMM-based algorithm for regu- larized SC. Experiments on various datasets demonstrate that weight vectors learned under our regularizer are more interpretable and have better generalization performance.
Tasks	Representation Learning
Published	2017-11-25
URL	http://arxiv.org/abs/1711.09300v1
PDF	http://arxiv.org/pdf/1711.09300v1.pdf
PWC	https://paperswithcode.com/paper/learning-less-overlapping-representations
Repo
Framework

Feature Squeezing Mitigates and Detects Carlini/Wagner Adversarial Examples


Title	Feature Squeezing Mitigates and Detects Carlini/Wagner Adversarial Examples
Authors	Weilin Xu, David Evans, Yanjun Qi
Abstract	Feature squeezing is a recently-introduced framework for mitigating and detecting adversarial examples. In previous work, we showed that it is effective against several earlier methods for generating adversarial examples. In this short note, we report on recent results showing that simple feature squeezing techniques also make deep learning models significantly more robust against the Carlini/Wagner attacks, which are the best known adversarial methods discovered to date.
Tasks
Published	2017-05-30
URL	http://arxiv.org/abs/1705.10686v1
PDF	http://arxiv.org/pdf/1705.10686v1.pdf
PWC	https://paperswithcode.com/paper/feature-squeezing-mitigates-and-detects
Repo
Framework

Bernoulli Rank-$1$ Bandits for Click Feedback


Title	Bernoulli Rank-$1$ Bandits for Click Feedback
Authors	Sumeet Katariya, Branislav Kveton, Csaba Szepesvári, Claire Vernade, Zheng Wen
Abstract	The probability that a user will click a search result depends both on its relevance and its position on the results page. The position based model explains this behavior by ascribing to every item an attraction probability, and to every position an examination probability. To be clicked, a result must be both attractive and examined. The probabilities of an item-position pair being clicked thus form the entries of a rank-$1$ matrix. We propose the learning problem of a Bernoulli rank-$1$ bandit where at each step, the learning agent chooses a pair of row and column arms, and receives the product of their Bernoulli-distributed values as a reward. This is a special case of the stochastic rank-$1$ bandit problem considered in recent work that proposed an elimination based algorithm Rank1Elim, and showed that Rank1Elim’s regret scales linearly with the number of rows and columns on “benign” instances. These are the instances where the minimum of the average row and column rewards $\mu$ is bounded away from zero. The issue with Rank1Elim is that it fails to be competitive with straightforward bandit strategies as $\mu \rightarrow 0$. In this paper we propose Rank1ElimKL which simply replaces the (crude) confidence intervals of Rank1Elim with confidence intervals based on Kullback-Leibler (KL) divergences, and with the help of a novel result concerning the scaling of KL divergences we prove that with this change, our algorithm will be competitive no matter the value of $\mu$. Experiments with synthetic data confirm that on benign instances the performance of Rank1ElimKL is significantly better than that of even Rank1Elim, while experiments with models derived from real data confirm that the improvements are significant across the board, regardless of whether the data is benign or not.
Tasks
Published	2017-03-19
URL	http://arxiv.org/abs/1703.06513v1
PDF	http://arxiv.org/pdf/1703.06513v1.pdf
PWC	https://paperswithcode.com/paper/bernoulli-rank-1-bandits-for-click-feedback
Repo
Framework

Finding Differentially Covarying Needles in a Temporally Evolving Haystack: A Scan Statistics Perspective


Title	Finding Differentially Covarying Needles in a Temporally Evolving Haystack: A Scan Statistics Perspective
Authors	Ronak Mehta, Hyunwoo J. Kim, Shulei Wang, Sterling C. Johnson, Ming Yuan, Vikas Singh
Abstract	Recent results in coupled or temporal graphical models offer schemes for estimating the relationship structure between features when the data come from related (but distinct) longitudinal sources. A novel application of these ideas is for analyzing group-level differences, i.e., in identifying if trends of estimated objects (e.g., covariance or precision matrices) are different across disparate conditions (e.g., gender or disease). Often, poor effect sizes make detecting the differential signal over the full set of features difficult: for example, dependencies between only a subset of features may manifest differently across groups. In this work, we first give a parametric model for estimating trends in the space of SPD matrices as a function of one or more covariates. We then generalize scan statistics to graph structures, to search over distinct subsets of features (graph partitions) whose temporal dependency structure may show statistically significant group-wise differences. We theoretically analyze the Family Wise Error Rate (FWER) and bounds on Type 1 and Type 2 error. On a cohort of individuals with risk factors for Alzheimer’s disease (but otherwise cognitively healthy), we find scientifically interesting group differences where the default analysis, i.e., models estimated on the full graph, do not survive reasonable significance thresholds.
Tasks
Published	2017-11-20
URL	http://arxiv.org/abs/1711.07575v1
PDF	http://arxiv.org/pdf/1711.07575v1.pdf
PWC	https://paperswithcode.com/paper/finding-differentially-covarying-needles-in-a
Repo
Framework

Synthetic Iris Presentation Attack using iDCGAN


Title	Synthetic Iris Presentation Attack using iDCGAN
Authors	Naman Kohli, Daksha Yadav, Mayank Vatsa, Richa Singh, Afzel Noore
Abstract	Reliability and accuracy of iris biometric modality has prompted its large-scale deployment for critical applications such as border control and national ID projects. The extensive growth of iris recognition systems has raised apprehensions about susceptibility of these systems to various attacks. In the past, researchers have examined the impact of various iris presentation attacks such as textured contact lenses and print attacks. In this research, we present a novel presentation attack using deep learning based synthetic iris generation. Utilizing the generative capability of deep convolutional generative adversarial networks and iris quality metrics, we propose a new framework, named as iDCGAN (iris deep convolutional generative adversarial network) for generating realistic appearing synthetic iris images. We demonstrate the effect of these synthetically generated iris images as presentation attack on iris recognition by using a commercial system. The state-of-the-art presentation attack detection framework, DESIST is utilized to analyze if it can discriminate these synthetically generated iris images from real images. The experimental results illustrate that mitigating the proposed synthetic presentation attack is of paramount importance.
Tasks	Iris Recognition
Published	2017-10-29
URL	http://arxiv.org/abs/1710.10565v1
PDF	http://arxiv.org/pdf/1710.10565v1.pdf
PWC	https://paperswithcode.com/paper/synthetic-iris-presentation-attack-using
Repo
Framework

Improving Chinese SRL with Heterogeneous Annotations


Title	Improving Chinese SRL with Heterogeneous Annotations
Authors	Qiaolin Xia, Baobao Chang, Zhifang Sui
Abstract	Previous studies on Chinese semantic role labeling (SRL) have concentrated on single semantically annotated corpus. But the training data of single corpus is often limited. Meanwhile, there usually exists other semantically annotated corpora for Chinese SRL scattered across different annotation frameworks. Data sparsity remains a bottleneck. This situation calls for larger training datasets, or effective approaches which can take advantage of highly heterogeneous data. In these papers, we focus mainly on the latter, that is, to improve Chinese SRL by using heterogeneous corpora together. We propose a novel progressive learning model which augments the Progressive Neural Network with Gated Recurrent Adapters. The model can accommodate heterogeneous inputs and effectively transfer knowledge between them. We also release a new corpus, Chinese SemBank, for Chinese SRL. Experiments on CPB 1.0 show that ours model outperforms state-of-the-art methods.
Tasks	Semantic Role Labeling
Published	2017-02-22
URL	http://arxiv.org/abs/1702.06740v3
PDF	http://arxiv.org/pdf/1702.06740v3.pdf
PWC	https://paperswithcode.com/paper/improving-chinese-srl-with-heterogeneous
Repo
Framework

LibOPT: An Open-Source Platform for Fast Prototyping Soft Optimization Techniques


Title	LibOPT: An Open-Source Platform for Fast Prototyping Soft Optimization Techniques
Authors	Joao Paulo Papa, Gustavo Henrique Rosa, Douglas Rodrigues, Xin-She Yang
Abstract	Optimization techniques play an important role in several scientific and real-world applications, thus becoming of great interest for the community. As a consequence, a number of open-source libraries are available in the literature, which ends up fostering the research and development of new techniques and applications. In this work, we present a new library for the implementation and fast prototyping of nature-inspired techniques called LibOPT. Currently, the library implements 15 techniques and 112 benchmarking functions, as well as it also supports 11 hypercomplex-based optimization approaches, which makes it one of the first of its kind. We showed how one can easily use and also implement new techniques in LibOPT under the C paradigm. Examples are provided with samples of source-code using benchmarking functions.
Tasks
Published	2017-04-18
URL	http://arxiv.org/abs/1704.05174v1
PDF	http://arxiv.org/pdf/1704.05174v1.pdf
PWC	https://paperswithcode.com/paper/libopt-an-open-source-platform-for-fast
Repo
Framework