May 5, 2019

3162 words 15 mins read

Paper Group ANR 485

High dimensional thresholded regression and shrinkage effect. Grid-like structure is optimal for path integration. Adaptive Training of Random Mapping for Data Quantization. Feature Learning from Spectrograms for Assessment of Personality Traits. A Novel Approach for Data-Driven Automatic Site Recommendation and Selection. Non-Central Catadioptric …

High dimensional thresholded regression and shrinkage effect


Title	High dimensional thresholded regression and shrinkage effect
Authors	Zemin Zheng, Yingying Fan, Jinchi Lv
Abstract	High-dimensional sparse modeling via regularization provides a powerful tool for analyzing large-scale data sets and obtaining meaningful, interpretable models. The use of nonconvex penalty functions shows advantage in selecting important features in high dimensions, but the global optimality of such methods still demands more understanding. In this paper, we consider sparse regression with hard-thresholding penalty, which we show to give rise to thresholded regression. This approach is motivated by its close connection with the $L_0$-regularization, which can be unrealistic to implement in practice but of appealing sampling properties, and its computational advantage. Under some mild regularity conditions allowing possibly exponentially growing dimensionality, we establish the oracle inequalities of the resulting regularized estimator, as the global minimizer, under various prediction and variable selection losses, as well as the oracle risk inequalities of the hard-thresholded estimator followed by a further $L_2$-regularization. The risk properties exhibit interesting shrinkage effects under both estimation and prediction losses. We identify the optimal choice of the ridge parameter, which is shown to have simultaneous advantages to both the $L_2$-loss and prediction loss. These new results and phenomena are evidenced by simulation and real data examples.
Tasks
Published	2016-05-11
URL	http://arxiv.org/abs/1605.03306v1
PDF	http://arxiv.org/pdf/1605.03306v1.pdf
PWC	https://paperswithcode.com/paper/high-dimensional-thresholded-regression-and
Repo
Framework

Grid-like structure is optimal for path integration


Title	Grid-like structure is optimal for path integration
Authors	Reza Moazzezi
Abstract	Grid cells in medial entorhinal cortex are believed to play a key role in path integration. However, the relation between path integration and the grid-like arrangement of their firing field remains unclear. We provide theoretical evidence that grid-like structure and path integration are closely related. In one dimension, the grid-like structure provides the optimal solution for path integration assuming that the noise correlation structure is Gaussian. In two dimensions, assuming that the noise is Gaussian, rectangular grid-like structure is the optimal solution provided that 1- both noise correlation and receptive field structures of the neurons can be multiplicatively decomposed into orthogonal components and 2- the eigenvalues of the decomposed correlation matrices decrease faster than the square of the frequency of the corresponding eigenvectors. We will also address the decoding mechanism and show that the problem of decoding reduces to the problem of extracting task relevant information in the presence of task irrelevant information. Change-based Population Coding provides the optimal solution for this problem.
Tasks
Published	2016-06-03
URL	http://arxiv.org/abs/1606.01239v1
PDF	http://arxiv.org/pdf/1606.01239v1.pdf
PWC	https://paperswithcode.com/paper/grid-like-structure-is-optimal-for-path
Repo
Framework

Adaptive Training of Random Mapping for Data Quantization


Title	Adaptive Training of Random Mapping for Data Quantization
Authors	Miao Cheng, Ah Chung Tsoi
Abstract	Data quantization learns encoding results of data with certain requirements, and provides a broad perspective of many real-world applications to data handling. Nevertheless, the results of encoder is usually limited to multivariate inputs with the random mapping, and side information of binary codes are hardly to mostly depict the original data patterns as possible. In the literature, cosine based random quantization has attracted much attentions due to its intrinsic bounded results. Nevertheless, it usually suffers from the uncertain outputs, and information of original data fails to be fully preserved in the reduced codes. In this work, a novel binary embedding method, termed adaptive training quantization (ATQ), is proposed to learn the ideal transform of random encoder, where the limitation of cosine random mapping is tackled. As an adaptive learning idea, the reduced mapping is adaptively calculated with idea of data group, while the bias of random transform is to be improved to hold most matching information. Experimental results show that the proposed method is able to obtain outstanding performance compared with other random quantization methods.
Tasks	Quantization
Published	2016-06-28
URL	http://arxiv.org/abs/1606.08808v2
PDF	http://arxiv.org/pdf/1606.08808v2.pdf
PWC	https://paperswithcode.com/paper/adaptive-training-of-random-mapping-for-data
Repo
Framework

Feature Learning from Spectrograms for Assessment of Personality Traits


Title	Feature Learning from Spectrograms for Assessment of Personality Traits
Authors	Marc-André Carbonneau, Eric Granger, Yazid Attabi, Ghyslain Gagnon
Abstract	Several methods have recently been proposed to analyze speech and automatically infer the personality of the speaker. These methods often rely on prosodic and other hand crafted speech processing features extracted with off-the-shelf toolboxes. To achieve high accuracy, numerous features are typically extracted using complex and highly parameterized algorithms. In this paper, a new method based on feature learning and spectrogram analysis is proposed to simplify the feature extraction process while maintaining a high level of accuracy. The proposed method learns a dictionary of discriminant features from patches extracted in the spectrogram representations of training speech segments. Each speech segment is then encoded using the dictionary, and the resulting feature set is used to perform classification of personality traits. Experiments indicate that the proposed method achieves state-of-the-art results with a significant reduction in complexity when compared to the most recent reference methods. The number of features, and difficulties linked to the feature extraction process are greatly reduced as only one type of descriptors is used, for which the 6 parameters can be tuned automatically. In contrast, the simplest reference method uses 4 types of descriptors to which 6 functionals are applied, resulting in over 20 parameters to be tuned.
Tasks
Published	2016-10-04
URL	http://arxiv.org/abs/1610.01223v1
PDF	http://arxiv.org/pdf/1610.01223v1.pdf
PWC	https://paperswithcode.com/paper/feature-learning-from-spectrograms-for
Repo
Framework

A Novel Approach for Data-Driven Automatic Site Recommendation and Selection


Title	A Novel Approach for Data-Driven Automatic Site Recommendation and Selection
Authors	Sebastian Baumbach, Frank Wittich, Florian Sachs, Sheraz Ahmed, Andreas Dengel
Abstract	This paper presents a novel, generic, and automatic method for data-driven site selection. Site selection is one of the most crucial and important decisions made by any company. Such a decision depends on various factors of sites, including socio-economic, geographical, ecological, as well as specific requirements of companies. The existing approaches for site selection (commonly used by economists) are manual, subjective, and not scalable, especially to Big Data. The presented method for site selection is robust, efficient, scalable, and is capable of handling challenges emerging in Big Data. To assess the effectiveness of the presented method, it is evaluated on real data (collected from Federal Statistical Office of Germany) of around 200 influencing factors which are considered by economists for site selection of Supermarkets in Germany (Lidl, EDEKA, and NP). Evaluation results show that there is a big overlap (86.4 %) between the sites of existing supermarkets and the sites recommended by the presented method. In addition, the method also recommends many sites (328) for supermarket where a store should be opened.
Tasks
Published	2016-08-03
URL	http://arxiv.org/abs/1608.01212v1
PDF	http://arxiv.org/pdf/1608.01212v1.pdf
PWC	https://paperswithcode.com/paper/a-novel-approach-for-data-driven-automatic
Repo
Framework

Non-Central Catadioptric Cameras Pose Estimation using 3D Lines


Title	Non-Central Catadioptric Cameras Pose Estimation using 3D Lines
Authors	Andre Mateus, Pedro Miraldo, Pedro U. Lima
Abstract	In this article we purpose a novel method for planar pose estimation of mobile robots. This method is based on an analytic solution (which we derived) for the projection of 3D straight lines, onto the mirror of Non-Central Catadioptric Cameras (NCCS). The resulting solution is rewritten as a function of the rotation and translation parameters, which is then used as an error function for a set of mirror points. Those should be the result of the projection of a set of points incident with the respective 3D lines. The camera’s pose is given by minimizing the error function, with the associated constraints. The method is validated by experiments both with synthetic and real data. The latter was collected from a mobile robot equipped with a NCCS.
Tasks	Pose Estimation
Published	2016-07-08
URL	http://arxiv.org/abs/1607.02290v1
PDF	http://arxiv.org/pdf/1607.02290v1.pdf
PWC	https://paperswithcode.com/paper/non-central-catadioptric-cameras-pose
Repo
Framework

On Unifying Multi-View Self-Representations for Clustering by Tensor Multi-Rank Minimization


Title	On Unifying Multi-View Self-Representations for Clustering by Tensor Multi-Rank Minimization
Authors	Yuan Xie, Dacheng Tao, Wensheng Zhang, Lei Zhang, Yan Liu, Yanyun Qu
Abstract	In this paper, we address the multi-view subspace clustering problem. Our method utilizes the circulant algebra for tensor, which is constructed by stacking the subspace representation matrices of different views and then rotating, to capture the low rank tensor subspace so that the refinement of the view-specific subspaces can be achieved, as well as the high order correlations underlying multi-view data can be explored.} By introducing a recently proposed tensor factorization, namely tensor-Singular Value Decomposition (t-SVD) \cite{kilmer13}, we can impose a new type of low-rank tensor constraint on the rotated tensor to capture the complementary information from multiple views. Different from traditional unfolding based tensor norm, this low-rank tensor constraint has optimality properties similar to that of matrix rank derived from SVD, so the complementary information among views can be explored more efficiently and thoroughly. The established model, called t-SVD based Multi-view Subspace Clustering (t-SVD-MSC), falls into the applicable scope of augmented Lagrangian method, and its minimization problem can be efficiently solved with theoretical convergence guarantee and relatively low computational complexity. Extensive experimental testing on eight challenging image dataset shows that the proposed method has achieved highly competent objective performance compared to several state-of-the-art multi-view clustering methods.
Tasks	Multi-view Subspace Clustering
Published	2016-10-23
URL	http://arxiv.org/abs/1610.07126v3
PDF	http://arxiv.org/pdf/1610.07126v3.pdf
PWC	https://paperswithcode.com/paper/on-unifying-multi-view-self-representations
Repo
Framework

Attention-based Memory Selection Recurrent Network for Language Modeling


Title	Attention-based Memory Selection Recurrent Network for Language Modeling
Authors	Da-Rong Liu, Shun-Po Chuang, Hung-yi Lee
Abstract	Recurrent neural networks (RNNs) have achieved great success in language modeling. However, since the RNNs have fixed size of memory, their memory cannot store all the information about the words it have seen before in the sentence, and thus the useful long-term information may be ignored when predicting the next words. In this paper, we propose Attention-based Memory Selection Recurrent Network (AMSRN), in which the model can review the information stored in the memory at each previous time step and select the relevant information to help generate the outputs. In AMSRN, the attention mechanism finds the time steps storing the relevant information in the memory, and memory selection determines which dimensions of the memory are involved in computing the attention weights and from which the information is extracted.In the experiments, AMSRN outperformed long short-term memory (LSTM) based language models on both English and Chinese corpora. Moreover, we investigate using entropy as a regularizer for attention weights and visualize how the attention mechanism helps language modeling.
Tasks	Language Modelling
Published	2016-11-26
URL	http://arxiv.org/abs/1611.08656v1
PDF	http://arxiv.org/pdf/1611.08656v1.pdf
PWC	https://paperswithcode.com/paper/attention-based-memory-selection-recurrent
Repo
Framework

Optically lightweight tracking of objects around a corner


Title	Optically lightweight tracking of objects around a corner
Authors	Jonathan Klein, Christoph Peters, Jaime Martín, Martin Laurenzis, Matthias B. Hullin
Abstract	The observation of objects located in inaccessible regions is a recurring challenge in a wide variety of important applications. Recent work has shown that indirect diffuse light reflections can be used to reconstruct objects and two-dimensional (2D) patterns around a corner. However, these prior methods always require some specialized setup involving either ultrafast detectors or narrowband light sources. Here we show that occluded objects can be tracked in real time using a standard 2D camera and a laser pointer. Unlike previous methods based on the backprojection approach, we formulate the problem in an analysis-by-synthesis sense. By repeatedly simulating light transport through the scene, we determine the set of object parameters that most closely fits the measured intensity distribution. We experimentally demonstrate that this approach is capable of following the translation of unknown objects, and translation and orientation of a known object, in real time.
Tasks
Published	2016-06-03
URL	http://arxiv.org/abs/1606.01873v1
PDF	http://arxiv.org/pdf/1606.01873v1.pdf
PWC	https://paperswithcode.com/paper/optically-lightweight-tracking-of-objects
Repo
Framework

Tensor-Dictionary Learning with Deep Kruskal-Factor Analysis


Title	Tensor-Dictionary Learning with Deep Kruskal-Factor Analysis
Authors	Andrew Stevens, Yunchen Pu, Yannan Sun, Greg Spell, Lawrence Carin
Abstract	A multi-way factor analysis model is introduced for tensor-variate data of any order. Each data item is represented as a (sparse) sum of Kruskal decompositions, a Kruskal-factor analysis (KFA). KFA is nonparametric and can infer both the tensor-rank of each dictionary atom and the number of dictionary atoms. The model is adapted for online learning, which allows dictionary learning on large data sets. After KFA is introduced, the model is extended to a deep convolutional tensor-factor analysis, supervised by a Bayesian SVM. The experiments section demonstrates the improvement of KFA over vectorized approaches (e.g., BPFA), tensor decompositions, and convolutional neural networks (CNN) in multi-way denoising, blind inpainting, and image classification. The improvement in PSNR for the inpainting results over other methods exceeds 1dB in several cases and we achieve state of the art results on Caltech101 image classification.
Tasks	Denoising, Dictionary Learning, Image Classification
Published	2016-12-08
URL	http://arxiv.org/abs/1612.02842v3
PDF	http://arxiv.org/pdf/1612.02842v3.pdf
PWC	https://paperswithcode.com/paper/tensor-dictionary-learning-with-deep-kruskal
Repo
Framework

Query-adaptive Image Retrieval by Deep Weighted Hashing


Title	Query-adaptive Image Retrieval by Deep Weighted Hashing
Authors	Jian Zhang, Yuxin Peng
Abstract	Hashing methods have attracted much attention for large scale image retrieval. Some deep hashing methods have achieved promising results by taking advantage of the strong representation power of deep networks recently. However, existing deep hashing methods treat all hash bits equally. On one hand, a large number of images share the same distance to a query image due to the discrete Hamming distance, which raises a critical issue of image retrieval where fine-grained rankings are very important. On the other hand, different hash bits actually contribute to the image retrieval differently, and treating them equally greatly affects the retrieval accuracy of image. To address the above two problems, we propose the query-adaptive deep weighted hashing (QaDWH) approach, which can perform fine-grained ranking for different queries by weighted Hamming distance. First, a novel deep hashing network is proposed to learn the hash codes and corresponding class-wise weights jointly, so that the learned weights can reflect the importance of different hash bits for different image classes. Second, a query-adaptive image retrieval method is proposed, which rapidly generates hash bit weights for different query images by fusing its semantic probability and the learned class-wise weights. Fine-grained image retrieval is then performed by the weighted Hamming distance, which can provide more accurate ranking than the traditional Hamming distance. Experiments on four widely used datasets show that the proposed approach outperforms eight state-of-the-art hashing methods.
Tasks	Image Retrieval
Published	2016-12-08
URL	http://arxiv.org/abs/1612.02541v2
PDF	http://arxiv.org/pdf/1612.02541v2.pdf
PWC	https://paperswithcode.com/paper/query-adaptive-image-retrieval-by-deep
Repo
Framework

Classification with Boosting of Extreme Learning Machine Over Arbitrarily Partitioned Data


Title	Classification with Boosting of Extreme Learning Machine Over Arbitrarily Partitioned Data
Authors	Ferhat Özgür Çatak
Abstract	Machine learning based computational intelligence methods are widely used to analyze large scale data sets in this age of big data. Extracting useful predictive modeling from these types of data sets is a challenging problem due to their high complexity. Analyzing large amount of streaming data that can be leveraged to derive business value is another complex problem to solve. With high levels of data availability (\textit{i.e. Big Data}) automatic classification of them has become an important and complex task. Hence, we explore the power of applying MapReduce based Distributed AdaBoosting of Extreme Learning Machine (ELM) to build a predictive bag of classification models. Accordingly, (i) data set ensembles are created; (ii) ELM algorithm is used to build weak learners (classifier functions); and (iii) builds a strong learner from a set of weak learners. We applied this training model to the benchmark knowledge discovery and data mining data sets.
Tasks
Published	2016-02-09
URL	http://arxiv.org/abs/1602.02887v1
PDF	http://arxiv.org/pdf/1602.02887v1.pdf
PWC	https://paperswithcode.com/paper/classification-with-boosting-of-extreme
Repo
Framework

Spot On: Action Localization from Pointly-Supervised Proposals


Title	Spot On: Action Localization from Pointly-Supervised Proposals
Authors	Pascal Mettes, Jan C. van Gemert, Cees G. M. Snoek
Abstract	We strive for spatio-temporal localization of actions in videos. The state-of-the-art relies on action proposals at test time and selects the best one with a classifier trained on carefully annotated box annotations. Annotating action boxes in video is cumbersome, tedious, and error prone. Rather than annotating boxes, we propose to annotate actions in video with points on a sparse subset of frames only. We introduce an overlap measure between action proposals and points and incorporate them all into the objective of a non-convex Multiple Instance Learning optimization. Experimental evaluation on the UCF Sports and UCF 101 datasets shows that (i) spatio-temporal proposals can be used to train classifiers while retaining the localization performance, (ii) point annotations yield results comparable to box annotations while being significantly faster to annotate, (iii) with a minimum amount of supervision our approach is competitive to the state-of-the-art. Finally, we introduce spatio-temporal action annotations on the train and test videos of Hollywood2, resulting in Hollywood2Tubes, available at http://tinyurl.com/hollywood2tubes.
Tasks	Action Localization, Multiple Instance Learning, Temporal Localization
Published	2016-04-26
URL	http://arxiv.org/abs/1604.07602v2
PDF	http://arxiv.org/pdf/1604.07602v2.pdf
PWC	https://paperswithcode.com/paper/spot-on-action-localization-from-pointly
Repo
Framework

Generalized Inverse Classification


Title	Generalized Inverse Classification
Authors	Michael T. Lash, Qihang Lin, W. Nick Street, Jennifer G. Robinson, Jeffrey Ohlmann
Abstract	Inverse classification is the process of perturbing an instance in a meaningful way such that it is more likely to conform to a specific class. Historical methods that address such a problem are often framed to leverage only a single classifier, or specific set of classifiers. These works are often accompanied by naive assumptions. In this work we propose generalized inverse classification (GIC), which avoids restricting the classification model that can be used. We incorporate this formulation into a refined framework in which GIC takes place. Under this framework, GIC operates on features that are immediately actionable. Each change incurs an individual cost, either linear or non-linear. Such changes are subjected to occur within a specified level of cumulative change (budget). Furthermore, our framework incorporates the estimation of features that change as a consequence of direct actions taken (indirectly changeable features). To solve such a problem, we propose three real-valued heuristic-based methods and two sensitivity analysis-based comparison methods, each of which is evaluated on two freely available real-world datasets. Our results demonstrate the validity and benefits of our formulation, framework, and methods.
Tasks
Published	2016-10-05
URL	http://arxiv.org/abs/1610.01675v2
PDF	http://arxiv.org/pdf/1610.01675v2.pdf
PWC	https://paperswithcode.com/paper/generalized-inverse-classification
Repo
Framework

Log-linear Combinations of Monolingual and Bilingual Neural Machine Translation Models for Automatic Post-Editing


Title	Log-linear Combinations of Monolingual and Bilingual Neural Machine Translation Models for Automatic Post-Editing
Authors	Marcin Junczys-Dowmunt, Roman Grundkiewicz
Abstract	This paper describes the submission of the AMU (Adam Mickiewicz University) team to the Automatic Post-Editing (APE) task of WMT 2016. We explore the application of neural translation models to the APE problem and achieve good results by treating different models as components in a log-linear model, allowing for multiple inputs (the MT-output and the source) that are decoded to the same target language (post-edited translations). A simple string-matching penalty integrated within the log-linear model is used to control for higher faithfulness with regard to the raw machine translation output. To overcome the problem of too little training data, we generate large amounts of artificial data. Our submission improves over the uncorrected baseline on the unseen test set by -3.2% TER and +5.5% BLEU and outperforms any other system submitted to the shared-task by a large margin.
Tasks	Automatic Post-Editing, Machine Translation
Published	2016-05-16
URL	http://arxiv.org/abs/1605.04800v2
PDF	http://arxiv.org/pdf/1605.04800v2.pdf
PWC	https://paperswithcode.com/paper/log-linear-combinations-of-monolingual-and
Repo
Framework