October 18, 2019

2861 words 14 mins read

Paper Group ANR 441

Heat Kernel analysis of Syntactic Structures. Depth Masked Discriminative Correlation Filter. Ordinal Regression using Noisy Pairwise Comparisons for Body Mass Index Range Estimation. End-to-end Learning of Convolutional Neural Net and Dynamic Programming for Left Ventricle Segmentation. Retinal Vessel Segmentation under Extreme Low Annotation: A G …

Heat Kernel analysis of Syntactic Structures


Title	Heat Kernel analysis of Syntactic Structures
Authors	Andrew Ortegaray, Robert C. Berwick, Matilde Marcolli
Abstract	We consider two different data sets of syntactic parameters and we discuss how to detect relations between parameters through a heat kernel method developed by Belkin-Niyogi, which produces low dimensional representations of the data, based on Laplace eigenfunctions, that preserve neighborhood information. We analyze the different connectivity and clustering structures that arise in the two datasets, and the regions of maximal variance in the two-parameter space of the Belkin-Niyogi construction, which identify preferable choices of independent variables. We compute clustering coefficients and their variance.
Tasks
Published	2018-03-26
URL	http://arxiv.org/abs/1803.09832v1
PDF	http://arxiv.org/pdf/1803.09832v1.pdf
PWC	https://paperswithcode.com/paper/heat-kernel-analysis-of-syntactic-structures
Repo
Framework

Depth Masked Discriminative Correlation Filter


Title	Depth Masked Discriminative Correlation Filter
Authors	Uğur Kart, Joni-Kristian Kämäräinen, Jiří Matas, Lixin Fan, Francesco Cricri
Abstract	Depth information provides a strong cue for occlusion detection and handling, but has been largely omitted in generic object tracking until recently due to lack of suitable benchmark datasets and applications. In this work, we propose a Depth Masked Discriminative Correlation Filter (DM-DCF) which adopts novel depth segmentation based occlusion detection that stops correlation filter updating and depth masking which adaptively adjusts the spatial support for correlation filter. In Princeton RGBD Tracking Benchmark, our DM-DCF is among the state-of-the-art in overall ranking and the winner on multiple categories. Moreover, since it is based on DCF, `DM-DCF` runs an order of magnitude faster than its competitors making it suitable for time constrained applications.
Tasks	Object Tracking
Published	2018-02-26
URL	http://arxiv.org/abs/1802.09227v2
PDF	http://arxiv.org/pdf/1802.09227v2.pdf
PWC	https://paperswithcode.com/paper/depth-masked-discriminative-correlation
Repo
Framework

Ordinal Regression using Noisy Pairwise Comparisons for Body Mass Index Range Estimation


Title	Ordinal Regression using Noisy Pairwise Comparisons for Body Mass Index Range Estimation
Authors	Luisa Polania, Dongning Wang, Glenn Fung
Abstract	Ordinal regression aims to classify instances into ordinal categories. In this paper, body mass index (BMI) category estimation from facial images is cast as an ordinal regression problem. In particular, noisy binary search algorithms based on pairwise comparisons are employed to exploit the ordinal relationship among BMI categories. Comparisons are performed with Siamese architectures, one of which uses the Bradley-Terry model probabilities as target. The Bradley-Terry model is an approach to describe probabilities of the possible outcomes when elements of a set are repeatedly compared with one another in pairs. Experimental results show that our approach outperforms classification and regression-based methods at estimating BMI categories.
Tasks
Published	2018-11-08
URL	http://arxiv.org/abs/1811.03268v1
PDF	http://arxiv.org/pdf/1811.03268v1.pdf
PWC	https://paperswithcode.com/paper/ordinal-regression-using-noisy-pairwise
Repo
Framework

End-to-end Learning of Convolutional Neural Net and Dynamic Programming for Left Ventricle Segmentation


Title	End-to-end Learning of Convolutional Neural Net and Dynamic Programming for Left Ventricle Segmentation
Authors	Nhat M. Nguyen, Nilanjan Ray
Abstract	Differentiable programming is able to combine different functions or programs in a processing pipeline with the goal of applying end-to-end learning or optimization. A significant impediment is the non-differentiable nature of some algorithms. We propose to use synthetic gradients (SG) to overcome this difficulty. SG uses the universal function approximation property of neural networks. We apply SG to combine convolutional neural network (CNN) with dynamic programming (DP) in end-to-end learning for segmenting left ventricle from short axis view of heart MRI. Our experiments show that end-to-end combination of CNN and DP requires fewer labeled images to achieve a significantly better segmentation accuracy than using only CNN.
Tasks
Published	2018-12-02
URL	https://arxiv.org/abs/1812.00328v2
PDF	https://arxiv.org/pdf/1812.00328v2.pdf
PWC	https://paperswithcode.com/paper/end-to-end-learning-of-convolutional-neural
Repo
Framework

Retinal Vessel Segmentation under Extreme Low Annotation: A Generative Adversarial Network Approach


Title	Retinal Vessel Segmentation under Extreme Low Annotation: A Generative Adversarial Network Approach
Authors	Avisek Lahiri, Vineet Jain, Arnab Mondal, Prabir Kumar Biswas
Abstract	Contemporary deep learning based medical image segmentation algorithms require hours of annotation labor by domain experts. These data hungry deep models perform sub-optimally in the presence of limited amount of labeled data. In this paper, we present a data efficient learning framework using the recent concept of Generative Adversarial Networks; this allows a deep neural network to perform significantly better than its fully supervised counterpart in low annotation regime. The proposed method is an extension of our previous work with the addition of a new unsupervised adversarial loss and a structured prediction based architecture. To the best of our knowledge, this work is the first demonstration of an adversarial framework based structured prediction model for medical image segmentation. Though generic, we apply our method for segmentation of blood vessels in retinal fundus images. We experiment with extreme low annotation budget (0.8 - 1.6% of contemporary annotation size). On DRIVE and STARE datasets, the proposed method outperforms our previous method and other fully supervised benchmark models by significant margins especially with very low number of annotated examples. In addition, our systematic ablation studies suggest some key recipes for successfully training GAN based semi-supervised algorithms with an encoder-decoder style network architecture.
Tasks	Medical Image Segmentation, Retinal Vessel Segmentation, Semantic Segmentation, Structured Prediction
Published	2018-09-05
URL	http://arxiv.org/abs/1809.01348v1
PDF	http://arxiv.org/pdf/1809.01348v1.pdf
PWC	https://paperswithcode.com/paper/retinal-vessel-segmentation-under-extreme-low
Repo
Framework

Few Sample Knowledge Distillation for Efficient Network Compression


Title	Few Sample Knowledge Distillation for Efficient Network Compression
Authors	Tianhong Li, Jianguo Li, Zhuang Liu, Changshui Zhang
Abstract	Deep neural network compression techniques such as pruning and weight tensor decomposition usually require fine-tuning to recover the prediction accuracy when the compression ratio is high. However, conventional fine-tuning suffers from the requirement of a large training set and the time-consuming training procedure. This paper proposes a novel solution for knowledge distillation from label-free few samples to realize both data efficiency and training/processing efficiency. We treat the original network as “teacher-net” and the compressed network as “student-net”. A 1x1 convolution layer is added at the end of each layer block of the student-net, and we fit the block-level outputs of the student-net to the teacher-net by estimating the parameters of the added layers. We prove that the added layer can be merged without adding extra parameters and computation cost during inference. Experiments on multiple datasets and network architectures verify the method’s effectiveness on student-nets obtained by various network pruning and weight decomposition methods. Our method can recover student-net’s accuracy to the same level as conventional fine-tuning methods in minutes while using only 1% label-free data of the full training data.
Tasks	Network Pruning, Neural Network Compression
Published	2018-12-05
URL	https://arxiv.org/abs/1812.01839v3
PDF	https://arxiv.org/pdf/1812.01839v3.pdf
PWC	https://paperswithcode.com/paper/knowledge-distillation-from-few-samples
Repo
Framework

Uncertainty-aware Short-term Motion Prediction of Traffic Actors for Autonomous Driving


Title	Uncertainty-aware Short-term Motion Prediction of Traffic Actors for Autonomous Driving
Authors	Nemanja Djuric, Vladan Radosavljevic, Henggang Cui, Thi Nguyen, Fang-Chieh Chou, Tsung-Han Lin, Nitin Singh, Jeff Schneider
Abstract	We address one of the crucial aspects necessary for safe and efficient operations of autonomous vehicles, namely predicting future state of traffic actors in the autonomous vehicle’s surroundings. We introduce a deep learning-based approach that takes into account a current world state and produces raster images of each actor’s vicinity. The rasters are then used as inputs to deep convolutional models to infer future movement of actors while also accounting for and capturing inherent uncertainty of the prediction task. Extensive experiments on real-world data strongly suggest benefits of the proposed approach. Moreover, following completion of the offline tests the system was successfully tested onboard self-driving vehicles.
Tasks	Autonomous Driving, Autonomous Vehicles, motion prediction
Published	2018-08-17
URL	https://arxiv.org/abs/1808.05819v3
PDF	https://arxiv.org/pdf/1808.05819v3.pdf
PWC	https://paperswithcode.com/paper/short-term-motion-prediction-of-traffic
Repo
Framework

META-DES.H: a dynamic ensemble selection technique using meta-learning and a dynamic weighting approach


Title	META-DES.H: a dynamic ensemble selection technique using meta-learning and a dynamic weighting approach
Authors	Rafael M. O. Cruz, Robert Sabourin, George D. C. Cavalcanti
Abstract	In Dynamic Ensemble Selection (DES) techniques, only the most competent classifiers are selected to classify a given query sample. Hence, the key issue in DES is how to estimate the competence of each classifier in a pool to select the most competent ones. In order to deal with this issue, we proposed a novel dynamic ensemble selection framework using meta-learning, called META-DES. The framework is divided into three steps. In the first step, the pool of classifiers is generated from the training data. In the second phase the meta-features are computed using the training data and used to train a meta-classifier that is able to predict whether or not a base classifier from the pool is competent enough to classify an input instance. In this paper, we propose improvements to the training and generalization phase of the META-DES framework. In the training phase, we evaluate four different algorithms for the training of the meta-classifier. For the generalization phase, three combination approaches are evaluated: Dynamic selection, where only the classifiers that attain a certain competence level are selected; Dynamic weighting, where the meta-classifier estimates the competence of each classifier in the pool, and the outputs of all classifiers in the pool are weighted based on their level of competence; and a hybrid approach, in which first an ensemble with the most competent classifiers is selected, after which the weights of the selected classifiers are estimated in order to be used in a weighted majority voting scheme. Experiments are carried out on 30 classification datasets. Experimental results demonstrate that the changes proposed in this paper significantly improve the recognition accuracy of the system in several datasets.
Tasks	Meta-Learning
Published	2018-11-01
URL	http://arxiv.org/abs/1811.01742v1
PDF	http://arxiv.org/pdf/1811.01742v1.pdf
PWC	https://paperswithcode.com/paper/meta-desh-a-dynamic-ensemble-selection
Repo
Framework

Cluster-based trajectory segmentation with local noise


Title	Cluster-based trajectory segmentation with local noise
Authors	Maria Luisa Damiani, Fatima Hachem, Issa Hamza, Nathan Ranc, Paul Moorcroft, Francesca Cagnacci
Abstract	We present a framework for the partitioning of a spatial trajectory in a sequence of segments based on spatial density and temporal criteria. The result is a set of temporally separated clusters interleaved by sub-sequences of unclustered points. A major novelty is the proposal of an outlier or noise model based on the distinction between intra-cluster (local noise) and inter-cluster noise (transition): the local noise models the temporary absence from a residence while the transition the definitive departure towards a next residence. We analyze in detail the properties of the model and present a comprehensive solution for the extraction of temporally ordered clusters. The effectiveness of the solution is evaluated first qualitatively and next quantitatively by contrasting the segmentation with ground truth. The ground truth consists of a set of trajectories of labeled points simulating animal movement. Moreover, we show that the approach can streamline the discovery of additional derived patterns, by presenting a novel technique for the analysis of periodic movement. From a methodological perspective, a valuable aspect of this research is that it combines the theoretical investigation with the application and external validation of the segmentation framework. This paves the way to an effective deployment of the solution in broad and challenging fields such as e-science.
Tasks
Published	2018-05-05
URL	http://arxiv.org/abs/1805.02102v1
PDF	http://arxiv.org/pdf/1805.02102v1.pdf
PWC	https://paperswithcode.com/paper/cluster-based-trajectory-segmentation-with
Repo
Framework

Robustifying Independent Component Analysis by Adjusting for Group-Wise Stationary Noise


Title	Robustifying Independent Component Analysis by Adjusting for Group-Wise Stationary Noise
Authors	Niklas Pfister, Sebastian Weichwald, Peter Bühlmann, Bernhard Schölkopf
Abstract	We introduce coroICA, confounding-robust independent component analysis, a novel ICA algorithm which decomposes linearly mixed multivariate observations into independent components that are corrupted (and rendered dependent) by hidden group-wise stationary confounding. It extends the ordinary ICA model in a theoretically sound and explicit way to incorporate group-wise (or environment-wise) confounding. We show that our proposed general noise model allows to perform ICA in settings where other noisy ICA procedures fail. Additionally, it can be used for applications with grouped data by adjusting for different stationary noise within each group. Our proposed noise model has a natural relation to causality and we explain how it can be applied in the context of causal inference. In addition to our theoretical framework, we provide an efficient estimation procedure and prove identifiability of the unmixing matrix under mild assumptions. Finally, we illustrate the performance and robustness of our method on simulated data, provide audible and visual examples, and demonstrate the applicability to real-world scenarios by experiments on publicly available Antarctic ice core data as well as two EEG data sets. We provide a scikit-learn compatible pip-installable Python package coroICA as well as R and Matlab implementations accompanied by a documentation at https://sweichwald.de/coroICA/
Tasks	Causal Inference, EEG
Published	2018-06-04
URL	https://arxiv.org/abs/1806.01094v3
PDF	https://arxiv.org/pdf/1806.01094v3.pdf
PWC	https://paperswithcode.com/paper/robustifying-independent-component-analysis
Repo
Framework

Unsupervised Learning of View-invariant Action Representations


Title	Unsupervised Learning of View-invariant Action Representations
Authors	Junnan Li, Yongkang Wong, Qi Zhao, Mohan S. Kankanhalli
Abstract	The recent success in human action recognition with deep learning methods mostly adopt the supervised learning paradigm, which requires significant amount of manually labeled data to achieve good performance. However, label collection is an expensive and time-consuming process. In this work, we propose an unsupervised learning framework, which exploits unlabeled data to learn video representations. Different from previous works in video representation learning, our unsupervised learning task is to predict 3D motion in multiple target views using video representation from a source view. By learning to extrapolate cross-view motions, the representation can capture view-invariant motion dynamics which is discriminative for the action. In addition, we propose a view-adversarial training method to enhance learning of view-invariant features. We demonstrate the effectiveness of the learned representations for action recognition on multiple datasets.
Tasks	Representation Learning, Temporal Action Localization
Published	2018-09-06
URL	http://arxiv.org/abs/1809.01844v1
PDF	http://arxiv.org/pdf/1809.01844v1.pdf
PWC	https://paperswithcode.com/paper/unsupervised-learning-of-view-invariant
Repo
Framework

A Discriminative Model for Identifying Readers and Assessing Text Comprehension from Eye Movements


Title	A Discriminative Model for Identifying Readers and Assessing Text Comprehension from Eye Movements
Authors	Silvia Makowski, Lena Jäger, Ahmed Abdelwahab, Niels Landwehr, Tobias Scheffer
Abstract	We study the problem of inferring readers’ identities and estimating their level of text comprehension from observations of their eye movements during reading. We develop a generative model of individual gaze patterns (scanpaths) that makes use of lexical features of the fixated words. Using this generative model, we derive a Fisher-score representation of eye-movement sequences. We study whether a Fisher-SVM with this Fisher kernel and several reference methods are able to identify readers and estimate their level of text comprehension based on eye-tracking data. While none of the methods are able to estimate text comprehension accurately, we find that the SVM with Fisher kernel excels at identifying readers.
Tasks	Eye Tracking, Reading Comprehension
Published	2018-09-21
URL	http://arxiv.org/abs/1809.08031v1
PDF	http://arxiv.org/pdf/1809.08031v1.pdf
PWC	https://paperswithcode.com/paper/a-discriminative-model-for-identifying
Repo
Framework

Predicting Periodicity with Temporal Difference Learning


Title	Predicting Periodicity with Temporal Difference Learning
Authors	Kristopher De Asis, Brendan Bennett, Richard S. Sutton
Abstract	Temporal difference (TD) learning is an important approach in reinforcement learning, as it combines ideas from dynamic programming and Monte Carlo methods in a way that allows for online and incremental model-free learning. A key idea of TD learning is that it is learning predictive knowledge about the environment in the form of value functions, from which it can derive its behavior to address long-term sequential decision making problems. The agent’s horizon of interest, that is, how immediate or long-term a TD learning agent predicts into the future, is adjusted through a discount rate parameter. In this paper, we introduce an alternative view on the discount rate, with insight from digital signal processing, to include complex-valued discounting. Our results show that setting the discount rate to appropriately chosen complex numbers allows for online and incremental estimation of the Discrete Fourier Transform (DFT) of a signal of interest with TD learning. We thereby extend the types of knowledge representable by value functions, which we show are particularly useful for identifying periodic effects in the reward sequence.
Tasks	Decision Making
Published	2018-09-20
URL	http://arxiv.org/abs/1809.07435v1
PDF	http://arxiv.org/pdf/1809.07435v1.pdf
PWC	https://paperswithcode.com/paper/predicting-periodicity-with-temporal
Repo
Framework

Mining useful Macro-actions in Planning


Title	Mining useful Macro-actions in Planning
Authors	Sandra Castellanos-Paez, Damien Pellier, Humbert Fiorino, Sylvie Pesty
Abstract	Planning has achieved significant progress in recent years. Among the various approaches to scale up plan synthesis, the use of macro-actions has been widely explored. As a first stage towards the development of a solution to learn on-line macro-actions, we propose an algorithm to identify useful macro-actions based on data mining techniques. The integration in the planning search of these learned macro-actions shows significant improvements over six classical planning benchmarks.
Tasks
Published	2018-10-22
URL	http://arxiv.org/abs/1810.09145v1
PDF	http://arxiv.org/pdf/1810.09145v1.pdf
PWC	https://paperswithcode.com/paper/mining-useful-macro-actions-in-planning
Repo
Framework

Fast On-the-fly Retraining-free Sparsification of Convolutional Neural Networks


Title	Fast On-the-fly Retraining-free Sparsification of Convolutional Neural Networks
Authors	Amir H. Ashouri, Tarek S. Abdelrahman, Alwyn Dos Remedios
Abstract	Modern Convolutional Neural Networks (CNNs) are complex, encompassing millions of parameters. Their deployment exerts computational, storage and energy demands, particularly on embedded platforms. Existing approaches to prune or sparsify CNNs require retraining to maintain inference accuracy. Such retraining is not feasible in some contexts. In this paper, we explore the sparsification of CNNs by proposing three model-independent methods. Our methods are applied on-the-fly and require no retraining. We show that the state-of-the-art models’ weights can be reduced by up to 73% (compression factor of 3.7x) without incurring more than 5% loss in Top-5 accuracy. Additional fine-tuning gains only 8% in sparsity, which indicates that our fast on-the-fly methods are effective.
Tasks
Published	2018-11-10
URL	https://arxiv.org/abs/1811.04199v3
PDF	https://arxiv.org/pdf/1811.04199v3.pdf
PWC	https://paperswithcode.com/paper/fast-on-the-fly-retraining-free
Repo
Framework