January 25, 2020

3020 words 15 mins read

Paper Group ANR 1627

Experience Reuse with Probabilistic Movement Primitives. Weakly Supervised Person Re-Identification. Emerging Cross-lingual Structure in Pretrained Language Models. Adaptive Gradient for Adversarial Perturbations Generation. Rethinking Convolutional Features in Correlation Filter Based Tracking. Measuring the influence of mere exposure effect of TV …

Experience Reuse with Probabilistic Movement Primitives


Title	Experience Reuse with Probabilistic Movement Primitives
Authors	Svenja Stark, Jan Peters, Elmar Rueckert
Abstract	Acquiring new robot motor skills is cumbersome, as learning a skill from scratch and without prior knowledge requires the exploration of a large space of motor configurations. Accordingly, for learning a new task, time could be saved by restricting the parameter search space by initializing it with the solution of a similar task. We present a framework which is able of such knowledge transfer from already learned movement skills to a new learning task. The framework combines probabilistic movement primitives with descriptions of their effects for skill representation. New skills are first initialized with parameters inferred from related movement primitives and thereafter adapted to the new task through relative entropy policy search. We compare two different transfer approaches to initialize the search space distribution with data of known skills with a similar effect. We show the different benefits of the two knowledge transfer approaches on an object pushing task for a simulated 3-DOF robot. We can show that the quality of the learned skills improves and the required iterations to learn a new task can be reduced by more than 60% when past experiences are utilized.
Tasks	Transfer Learning
Published	2019-08-11
URL	https://arxiv.org/abs/1908.03936v2
PDF	https://arxiv.org/pdf/1908.03936v2.pdf
PWC	https://paperswithcode.com/paper/experience-reuse-with-probabilistic-movement
Repo
Framework

Weakly Supervised Person Re-Identification


Title	Weakly Supervised Person Re-Identification
Authors	Jingke Meng, Sheng Wu, Wei-Shi Zheng
Abstract	In the conventional person re-id setting, it is assumed that the labeled images are the person images within the bounding box for each individual; this labeling across multiple nonoverlapping camera views from raw video surveillance is costly and time-consuming. To overcome this difficulty, we consider weakly supervised person re-id modeling. The weak setting refers to matching a target person with an untrimmed gallery video where we only know that the identity appears in the video without the requirement of annotating the identity in any frame of the video during the training procedure. Hence, for a video, there could be multiple video-level labels. We cast this weakly supervised person re-id challenge into a multi-instance multi-label learning (MIML) problem. In particular, we develop a Cross-View MIML (CV-MIML) method that is able to explore potential intraclass person images from all the camera views by incorporating the intra-bag alignment and the cross-view bag alignment. Finally, the CV-MIML method is embedded into an existing deep neural network for developing the Deep Cross-View MIML (Deep CV-MIML) model. We have performed extensive experiments to show the feasibility of the proposed weakly supervised setting and verify the effectiveness of our method compared to related methods on four weakly labeled datasets.
Tasks	Multi-Label Learning, Person Re-Identification
Published	2019-04-08
URL	https://arxiv.org/abs/1904.03832v2
PDF	https://arxiv.org/pdf/1904.03832v2.pdf
PWC	https://paperswithcode.com/paper/weakly-supervised-person-re-identification-1
Repo
Framework

Emerging Cross-lingual Structure in Pretrained Language Models


Title	Emerging Cross-lingual Structure in Pretrained Language Models
Authors	Shijie Wu, Alexis Conneau, Haoran Li, Luke Zettlemoyer, Veselin Stoyanov
Abstract	We study the problem of multilingual masked language modeling, i.e. the training of a single model on concatenated text from multiple languages, and present a detailed study of several factors that influence why these models are so effective for cross-lingual transfer. We show, contrary to what was previously hypothesized, that transfer is possible even when there is no shared vocabulary across the monolingual corpora and also when the text comes from very different domains. The only requirement is that there are some shared parameters in the top layers of the multi-lingual encoder. To better understand this result, we also show that representations from independently trained models in different languages can be aligned post-hoc quite effectively, strongly suggesting that, much like for non-contextual word embeddings, there are universal latent symmetries in the learned embedding spaces. For multilingual masked language modeling, these symmetries seem to be automatically discovered and aligned during the joint training process.
Tasks	Cross-Lingual Transfer, Language Modelling, Word Embeddings
Published	2019-11-04
URL	https://arxiv.org/abs/1911.01464v2
PDF	https://arxiv.org/pdf/1911.01464v2.pdf
PWC	https://paperswithcode.com/paper/emerging-cross-lingual-structure-in
Repo
Framework

Adaptive Gradient for Adversarial Perturbations Generation


Title	Adaptive Gradient for Adversarial Perturbations Generation
Authors	Yatie Xiao, Chi-Man Pun
Abstract	Deep Neural Networks have achieved remarkable success in computer vision, natural language processing, and audio tasks.
Tasks	Image Classification
Published	2019-02-01
URL	https://arxiv.org/abs/1902.01220v6
PDF	https://arxiv.org/pdf/1902.01220v6.pdf
PWC	https://paperswithcode.com/paper/adversarial-example-generation
Repo
Framework

Rethinking Convolutional Features in Correlation Filter Based Tracking


Title	Rethinking Convolutional Features in Correlation Filter Based Tracking
Authors	Fang Liang, Wenjun Peng, Qinghao Liu, Haijin Wang
Abstract	Both accuracy and efficiency are of significant importance to the task of visual object tracking. In recent years, as the surge of deep learning, Deep Convolutional NeuralNetwork (DCNN) becomes a very popular choice among the tracking community. However, due to the high computational complexity, end-to-end visual object trackers can hardly achieve an acceptable inference time and therefore can difficult to be utilized in many real-world applications. In this paper, we revisit a hierarchical deep feature-based visual tracker and found that both the performance and efficiency of the deep tracker are limited by the poor feature quality. Therefore, we propose a feature selection module to select more discriminative features for the trackers. After removing redundant features, our proposed tracker achieves significant improvements in both performance and efficiency. Finally, comparisons with state-of-the-art trackers are provided.
Tasks	Feature Selection, Object Tracking, Visual Object Tracking
Published	2019-12-30
URL	https://arxiv.org/abs/1912.12811v1
PDF	https://arxiv.org/pdf/1912.12811v1.pdf
PWC	https://paperswithcode.com/paper/rethinking-convolutional-features-in
Repo
Framework

Measuring the influence of mere exposure effect of TV commercial adverts on purchase behavior based on machine learning prediction models


Title	Measuring the influence of mere exposure effect of TV commercial adverts on purchase behavior based on machine learning prediction models
Authors	Elisa Claire Alemán Carreón, Hirofumi Nonaka, Asahi Hentona, Hirochika Yamashiro
Abstract	Since its introduction, television has been the main channel of investment for advertisements in order to influence customers purchase behavior. Many have attributed the mere exposure effect as the source of influence in purchase intention and purchase decision; however, most of the studies of television advertisement effects are not only outdated, but their sample size is questionable and their environments do not reflect reality. With the advent of the internet, social media and new information technologies, many recent studies focus on the effects of online advertisement, meanwhile, the investment in television advertisement still has not declined. In response to this, we applied machine learning algorithms SVM and XGBoost, as well as Logistic Regression, to construct a number of prediction models based on at-home advertisement exposure time and demographic data, examining the predictability of Actual Purchase and Purchase Intention behaviors of 3000 customers across 36 different products during the span of 3 months. If models based on exposure time had unreliable predictability in contrast to models based on demographic data, doubts would surface about the effectiveness of the hard investment in television advertising. Based on our results, we found that models based on advert exposure time were consistently low in their predictability in comparison with models based on demographic data only, and with models based on both demographic data and exposure time data. We also found that there was not a statistically significant difference between these last two kinds of models. This suggests that advert exposure time has little to no effect in the short-term in increasing positive actual purchase behavior.
Tasks
Published	2019-04-15
URL	http://arxiv.org/abs/1904.06862v2
PDF	http://arxiv.org/pdf/1904.06862v2.pdf
PWC	https://paperswithcode.com/paper/measuring-the-influence-of-mere-exposure
Repo
Framework

Patch alignment manifold matting


Title	Patch alignment manifold matting
Authors	Xuelong Li, Kang Liu, Yongsheng Dong, Dacheng Tao
Abstract	Image matting is generally modeled as a space transform from the color space to the alpha space. By estimating the alpha factor of the model, the foreground of an image can be extracted. However, there is some dimensional information redundancy in the alpha space. It usually leads to the misjudgments of some pixels near the boundary between the foreground and the background. In this paper, a manifold matting framework named Patch Alignment Manifold Matting is proposed for image matting. In particular, we first propose a part modeling of color space in the local image patch. We then perform whole alignment optimization for approximating the alpha results using subspace reconstructing error. Furthermore, we utilize Nesterov’s algorithm to solve the optimization problem. Finally, we apply some manifold learning methods in the framework, and obtain several image matting methods, such as named ISOMAP matting and its derived Cascade ISOMAP matting. The experimental results reveal that the manifold matting framework and its two examples are effective when compared with several representative matting methods.
Tasks	Image Matting
Published	2019-04-16
URL	http://arxiv.org/abs/1904.07588v1
PDF	http://arxiv.org/pdf/1904.07588v1.pdf
PWC	https://paperswithcode.com/paper/patch-alignment-manifold-matting
Repo
Framework

Progress Notes Classification and Keyword Extraction using Attention-based Deep Learning Models with BERT


Title	Progress Notes Classification and Keyword Extraction using Attention-based Deep Learning Models with BERT
Authors	Matthew Tang, Priyanka Gandhi, Md Ahsanul Kabir, Christopher Zou, Jordyn Blakey, Xiao Luo
Abstract	Various deep learning algorithms have been developed to analyze different types of clinical data including clinical text classification and extracting information from ‘free text’ and so on. However, automate the keyword extraction from the clinical notes is still challenging. The challenges include dealing with noisy clinical notes which contain various abbreviations, possible typos, and unstructured sentences. The objective of this research is to investigate the attention-based deep learning models to classify the de-identified clinical progress notes extracted from a real-world EHR system. The attention-based deep learning models can be used to interpret the models and understand the critical words that drive the correct or incorrect classification of the clinical progress notes. The attention-based models in this research are capable of presenting the human interpretable text classification models. The results show that the fine-tuned BERT with the attention layer can achieve a high classification accuracy of 97.6%, which is higher than the baseline fine-tuned BERT classification model. In this research, we also demonstrate that the attention-based models can identify relevant keywords that are strongly related to the clinical progress note categories.
Tasks	Document Classification, Keyword Extraction, Text Classification
Published	2019-10-13
URL	https://arxiv.org/abs/1910.05786v2
PDF	https://arxiv.org/pdf/1910.05786v2.pdf
PWC	https://paperswithcode.com/paper/progress-notes-classification-and-keyword
Repo
Framework

Analyzing symmetry and symmetry breaking by computational aesthetic measures


Title	Analyzing symmetry and symmetry breaking by computational aesthetic measures
Authors	Hendrik Richter
Abstract	We study creating and analyzing symmetry and broken symmetry in digital art. Our focus is not so much on computer-generating artistic images, but rather on analyzing concepts and templates for incorporating symmetry and symmetry breaking into the creation process. Taking as a starting point patterns generated algorithmically by emulating the collective feeding behavior of sand-bubbler crabs, all four types of two-dimensional symmetry are used as isometric maps. Apart from a geometric interpretation of symmetry, we also consider color as an object of symmetric transformations. Color symmetry is realized as a color permutation consistent with the isometries. Moreover, we analyze the abilities of computational aesthetic measures to serve as a metric that reflects design parameters, i.e. the type of symmetry and the degree of symmetry breaking.
Tasks
Published	2019-10-15
URL	https://arxiv.org/abs/1910.06610v1
PDF	https://arxiv.org/pdf/1910.06610v1.pdf
PWC	https://paperswithcode.com/paper/analyzing-symmetry-and-symmetry-breaking-by
Repo
Framework

Long-Term Vehicle Localization by Recursive Knowledge Distillation


Title	Long-Term Vehicle Localization by Recursive Knowledge Distillation
Authors	Hiroki Tomoe, Tanaka Kanji
Abstract	Most of the current state-of-the-art frameworks for cross-season visual place recognition (CS-VPR) focus on domain adaptation (DA) to a single specific season. From the viewpoint of long-term CS-VPR, such frameworks do not scale well to sequential multiple domains (e.g., spring - summer - autumn - winter - … ). The goal of this study is to develop a novel long-term ensemble learning (LEL) framework that allows for a constant cost retraining in long-term sequential-multi-domain CS-VPR (SMD-VPR), which only requires the memorization of a small constant number of deep convolutional neural networks (CNNs) and can retrain the CNN ensemble of every season at a small constant time/space cost. We frame our task as the multi-teacher multi-student knowledge distillation (MTMS-KD), which recursively compresses all the previous season’s knowledge into a current CNN ensemble. We further address the issue of teacher-student-assignment (TSA) to achieve a good generalization/specialization tradeoff. Experimental results on SMD-VPR tasks validate the efficacy of the proposed approach.
Tasks	Domain Adaptation, Visual Place Recognition
Published	2019-04-07
URL	http://arxiv.org/abs/1904.03551v1
PDF	http://arxiv.org/pdf/1904.03551v1.pdf
PWC	https://paperswithcode.com/paper/long-term-vehicle-localization-by-recursive
Repo
Framework

The MBPEP: a deep ensemble pruning algorithm providing high quality uncertainty prediction


Title	The MBPEP: a deep ensemble pruning algorithm providing high quality uncertainty prediction
Authors	Ruihan Hu, Qijun Huang, Sheng Chang, Hao Wang, Jin He
Abstract	Machine learning algorithms have been effectively applied into various real world tasks. However, it is difficult to provide high-quality machine learning solutions to accommodate an unknown distribution of input datasets; this difficulty is called the uncertainty prediction problems. In this paper, a margin-based Pareto deep ensemble pruning (MBPEP) model is proposed. It achieves the high-quality uncertainty estimation with a small value of the prediction interval width (MPIW) and a high confidence of prediction interval coverage probability (PICP) by using deep ensemble networks. In addition to these networks, unique loss functions are proposed, and these functions make the sub-learners available for standard gradient descent learning. Furthermore, the margin criterion fine-tuning-based Pareto pruning method is introduced to optimize the ensembles. Several experiments including predicting uncertainties of classification and regression are conducted to analyze the performance of MBPEP. The experimental results show that MBPEP achieves a small interval width and a low learning error with an optimal number of ensembles. For the real-world problems, MBPEP performs well on input datasets with unknown distributions datasets incomings and improves learning performance on a multi task problem when compared to that of each single model.
Tasks
Published	2019-02-25
URL	http://arxiv.org/abs/1902.09238v1
PDF	http://arxiv.org/pdf/1902.09238v1.pdf
PWC	https://paperswithcode.com/paper/the-mbpep-a-deep-ensemble-pruning-algorithm
Repo
Framework

Adversarial Exploitation of Policy Imitation


Title	Adversarial Exploitation of Policy Imitation
Authors	Vahid Behzadan, William Hsu
Abstract	This paper investigates a class of attacks targeting the confidentiality aspect of security in Deep Reinforcement Learning (DRL) policies. Recent research have established the vulnerability of supervised machine learning models (e.g., classifiers) to model extraction attacks. Such attacks leverage the loosely-restricted ability of the attacker to iteratively query the model for labels, thereby allowing for the forging of a labeled dataset which can be used to train a replica of the original model. In this work, we demonstrate the feasibility of exploiting imitation learning techniques in launching model extraction attacks on DRL agents. Furthermore, we develop proof-of-concept attacks that leverage such techniques for black-box attacks against the integrity of DRL policies. We also present a discussion on potential solution concepts for mitigation techniques.
Tasks	Imitation Learning
Published	2019-06-03
URL	https://arxiv.org/abs/1906.01121v1
PDF	https://arxiv.org/pdf/1906.01121v1.pdf
PWC	https://paperswithcode.com/paper/adversarial-exploitation-of-policy-imitation
Repo
Framework

The Stanford Acuity Test: A Precise Vision Test Using Bayesian Techniques and a Discovery in Human Visual Response


Title	The Stanford Acuity Test: A Precise Vision Test Using Bayesian Techniques and a Discovery in Human Visual Response
Authors	Chris Piech, Ali Malik, Laura M Scott, Robert T Chang, Charles Lin
Abstract	Chart-based visual acuity measurements are used by billions of people to diagnose and guide treatment of vision impairment. However, the ubiquitous eye exam has no mechanism for reasoning about uncertainty and as such, suffers from a well-documented reproducibility problem. In this paper we make two core contributions. First, we uncover a new parametric probabilistic model of visual acuity response based on detailed measurements of patients with eye disease. Then, we present an adaptive, digital eye exam using modern artificial intelligence techniques which substantially reduces acuity exam error over existing approaches, while also introducing the novel ability to model its own uncertainty and incorporate prior beliefs. Using standard evaluation metrics, we estimate a 74% reduction in prediction error compared to the ubiquitous chart-based eye exam and up to 67% reduction compared to the previous best digital exam. For patients with eye disease, the novel ability to finely measure acuity from home could be a crucial part in early diagnosis. We provide a web implementation of our algorithm for anyone in the world to use. The insights in this paper also provide interesting implications for the field of psychometric Item Response Theory.
Tasks
Published	2019-06-05
URL	https://arxiv.org/abs/1906.01811v2
PDF	https://arxiv.org/pdf/1906.01811v2.pdf
PWC	https://paperswithcode.com/paper/the-stanford-acuity-test-a-probabilistic
Repo
Framework

Compact Scene Graphs for Layout Composition and Patch Retrieval


Title	Compact Scene Graphs for Layout Composition and Patch Retrieval
Authors	Subarna Tripathi, Sharath Nittur Sridhar, Sairam Sundaresan, Hanlin Tang
Abstract	Structured representations such as scene graphs serve as an efficient and compact representation that can be used for downstream rendering or retrieval tasks. However, existing efforts to generate realistic images from scene graphs perform poorly on scene composition for cluttered or complex scenes. We propose two contributions to improve the scene composition. First, we enhance the scene graph representation with heuristic-based relations, which add minimal storage overhead. Second, we use extreme points representation to supervise the learning of the scene composition network. These methods achieve significantly higher performance over existing work (69.0% vs 51.2% in relation score metric). We additionally demonstrate how scene graphs can be used to retrieve pose-constrained image patches that are semantically similar to the source query. Improving structured scene graph representations for rendering or retrieval is an important step towards realistic image generation.
Tasks	Image Generation
Published	2019-04-19
URL	http://arxiv.org/abs/1904.09348v1
PDF	http://arxiv.org/pdf/1904.09348v1.pdf
PWC	https://paperswithcode.com/paper/190409348
Repo
Framework

Simple Automatic Post-editing for Arabic-Japanese Machine Translation


Title	Simple Automatic Post-editing for Arabic-Japanese Machine Translation
Authors	Ella Noll, Mai Oudah, Nizar Habash
Abstract	A common bottleneck for developing machine translation (MT) systems for some language pairs is the lack of direct parallel translation data sets, in general and in certain domains. Alternative solutions such as zero-shot models or pivoting techniques are successful in getting a strong baseline, but are often below the more supported language-pair systems. In this paper, we focus on Arabic-Japanese machine translation, a less studied language pair; and we work with a unique parallel corpus of Arabic news articles that were manually translated to Japanese. We use this parallel corpus to adapt a state-of-the-art domain/genre agnostic neural MT system via a simple automatic post-editing technique. Our results and detailed analysis suggest that this approach is quite viable for less supported language pairs in specific domains.
Tasks	Automatic Post-Editing, Machine Translation
Published	2019-07-14
URL	https://arxiv.org/abs/1907.06210v1
PDF	https://arxiv.org/pdf/1907.06210v1.pdf
PWC	https://paperswithcode.com/paper/simple-automatic-post-editing-for-arabic
Repo
Framework