July 27, 2019

2805 words 14 mins read

Paper Group ANR 567

Aperture Supervision for Monocular Depth Estimation. A reinforcement learning algorithm for building collaboration in multi-agent systems. Serious Games Application for Memory Training Using Egocentric Images. SafePredict: A Meta-Algorithm for Machine Learning That Uses Refusals to Guarantee Correctness. Object Classification in Images of Neoclassi …

Aperture Supervision for Monocular Depth Estimation


Title	Aperture Supervision for Monocular Depth Estimation
Authors	Pratul P. Srinivasan, Rahul Garg, Neal Wadhwa, Ren Ng, Jonathan T. Barron
Abstract	We present a novel method to train machine learning algorithms to estimate scene depths from a single image, by using the information provided by a camera’s aperture as supervision. Prior works use a depth sensor’s outputs or images of the same scene from alternate viewpoints as supervision, while our method instead uses images from the same viewpoint taken with a varying camera aperture. To enable learning algorithms to use aperture effects as supervision, we introduce two differentiable aperture rendering functions that use the input image and predicted depths to simulate the depth-of-field effects caused by real camera apertures. We train a monocular depth estimation network end-to-end to predict the scene depths that best explain these finite aperture images as defocus-blurred renderings of the input all-in-focus image.
Tasks	Depth Estimation, Monocular Depth Estimation
Published	2017-11-21
URL	http://arxiv.org/abs/1711.07933v2
PDF	http://arxiv.org/pdf/1711.07933v2.pdf
PWC	https://paperswithcode.com/paper/aperture-supervision-for-monocular-depth
Repo
Framework

A reinforcement learning algorithm for building collaboration in multi-agent systems


Title	A reinforcement learning algorithm for building collaboration in multi-agent systems
Authors	Mehmet Emin Aydin, Ryan Fellows
Abstract	This paper presents a proof-of concept study for demonstrating the viability of building collaboration among multiple agents through standard Q learning algorithm embedded in particle swarm optimisation. Collaboration is formulated to be achieved among the agents via some sort competition, where the agents are expected to balance their action in such a way that none of them drifts away of the team and none intervene any fellow neighbours territory. Particles are devised with Q learning algorithm for self training to learn how to act as members of a swarm and how to produce collaborative/collective behaviours. The produced results are supportive to the algorithmic structures suggesting that a substantive collaboration can be build via proposed learning algorithm.
Tasks	Q-Learning
Published	2017-11-28
URL	http://arxiv.org/abs/1711.10574v2
PDF	http://arxiv.org/pdf/1711.10574v2.pdf
PWC	https://paperswithcode.com/paper/a-reinforcement-learning-algorithm-for
Repo
Framework

Serious Games Application for Memory Training Using Egocentric Images


Title	Serious Games Application for Memory Training Using Egocentric Images
Authors	Gabriel Oliveira-Barra, Marc Bolaños, Estefania Talavera, Adrián Dueñas, Olga Gelonch, Maite Garolera
Abstract	Mild cognitive impairment is the early stage of several neurodegenerative diseases, such as Alzheimer’s. In this work, we address the use of lifelogging as a tool to obtain pictures from a patient’s daily life from an egocentric point of view. We propose to use them in combination with serious games as a way to provide a non-pharmacological treatment to improve their quality of life. To do so, we introduce a novel computer vision technique that classifies rich and non rich egocentric images and uses them in serious games. We present results over a dataset composed by 10,997 images, recorded by 7 different users, achieving 79% of F1-score. Our model presents the first method used for automatic egocentric images selection applicable to serious games.
Tasks
Published	2017-07-27
URL	http://arxiv.org/abs/1707.08821v1
PDF	http://arxiv.org/pdf/1707.08821v1.pdf
PWC	https://paperswithcode.com/paper/serious-games-application-for-memory-training
Repo
Framework

SafePredict: A Meta-Algorithm for Machine Learning That Uses Refusals to Guarantee Correctness


Title	SafePredict: A Meta-Algorithm for Machine Learning That Uses Refusals to Guarantee Correctness
Authors	Mustafa A. Kocak, David Ramirez, Elza Erkip, Dennis E. Shasha
Abstract	SafePredict is a novel meta-algorithm that works with any base prediction algorithm for online data to guarantee an arbitrarily chosen correctness rate, $1-\epsilon$, by allowing refusals. Allowing refusals means that the meta-algorithm may refuse to emit a prediction produced by the base algorithm on occasion so that the error rate on non-refused predictions does not exceed $\epsilon$. The SafePredict error bound does not rely on any assumptions on the data distribution or the base predictor. When the base predictor happens not to exceed the target error rate $\epsilon$, SafePredict refuses only a finite number of times. When the error rate of the base predictor changes through time SafePredict makes use of a weight-shifting heuristic that adapts to these changes without knowing when the changes occur yet still maintains the correctness guarantee. Empirical results show that (i) SafePredict compares favorably with state-of-the art confidence based refusal mechanisms which fail to offer robust error guarantees; and (ii) combining SafePredict with such refusal mechanisms can in many cases further reduce the number of refusals. Our software (currently in Python) is included in the supplementary material.
Tasks
Published	2017-08-21
URL	http://arxiv.org/abs/1708.06425v2
PDF	http://arxiv.org/pdf/1708.06425v2.pdf
PWC	https://paperswithcode.com/paper/safepredict-a-meta-algorithm-for-machine
Repo
Framework

Object Classification in Images of Neoclassical Artifacts Using Deep Learning


Title	Object Classification in Images of Neoclassical Artifacts Using Deep Learning
Authors	Bernhard Bermeitinger, Maria Christoforaki, Simon Donig, Siegfried Handschuh
Abstract	In this paper, we report on our efforts for using Deep Learning for classifying artifacts and their features in digital visuals as a part of the Neoclassica framework. It was conceived to provide scholars with new methods for analyzing and classifying artifacts and aesthetic forms from the era of Classicism. The framework accommodates both traditional knowledge representation as a formal ontology and data-driven knowledge discovery, where cultural patterns will be identified by means of algorithms in statistical analysis and machine learning. We created a Deep Learning approach trained on photographs to classify the objects inside these photographs. In a next step, we will apply a different Deep Learning approach. It is capable of locating multiple objects inside an image and classifying them with a high accuracy.
Tasks	Object Classification
Published	2017-10-13
URL	http://arxiv.org/abs/1710.04943v1
PDF	http://arxiv.org/pdf/1710.04943v1.pdf
PWC	https://paperswithcode.com/paper/object-classification-in-images-of
Repo
Framework

Safer Classification by Synthesis


Title	Safer Classification by Synthesis
Authors	William Wang, Angelina Wang, Aviv Tamar, Xi Chen, Pieter Abbeel
Abstract	The discriminative approach to classification using deep neural networks has become the de-facto standard in various fields. Complementing recent reservations about safety against adversarial examples, we show that conventional discriminative methods can easily be fooled to provide incorrect labels with very high confidence to out of distribution examples. We posit that a generative approach is the natural remedy for this problem, and propose a method for classification using generative models. At training time, we learn a generative model for each class, while at test time, given an example to classify, we query each generator for its most similar generation, and select the class corresponding to the most similar one. Our approach is general and can be used with expressive models such as GANs and VAEs. At test time, our method accurately “knows when it does not know,” and provides resilience to out of distribution examples while maintaining competitive performance for standard examples.
Tasks
Published	2017-11-22
URL	http://arxiv.org/abs/1711.08534v2
PDF	http://arxiv.org/pdf/1711.08534v2.pdf
PWC	https://paperswithcode.com/paper/safer-classification-by-synthesis
Repo
Framework

Skeleton-based Action Recognition with Convolutional Neural Networks


Title	Skeleton-based Action Recognition with Convolutional Neural Networks
Authors	Chao Li, Qiaoyong Zhong, Di Xie, Shiliang Pu
Abstract	Current state-of-the-art approaches to skeleton-based action recognition are mostly based on recurrent neural networks (RNN). In this paper, we propose a novel convolutional neural networks (CNN) based framework for both action classification and detection. Raw skeleton coordinates as well as skeleton motion are fed directly into CNN for label prediction. A novel skeleton transformer module is designed to rearrange and select important skeleton joints automatically. With a simple 7-layer network, we obtain 89.3% accuracy on validation set of the NTU RGB+D dataset. For action detection in untrimmed videos, we develop a window proposal network to extract temporal segment proposals, which are further classified within the same network. On the recent PKU-MMD dataset, we achieve 93.7% mAP, surpassing the baseline by a large margin.
Tasks	Action Classification, Action Detection, Skeleton Based Action Recognition, Temporal Action Localization
Published	2017-04-25
URL	http://arxiv.org/abs/1704.07595v1
PDF	http://arxiv.org/pdf/1704.07595v1.pdf
PWC	https://paperswithcode.com/paper/skeleton-based-action-recognition-with-2
Repo
Framework

Segmented and Non-Segmented Stacked Denoising Autoencoder for Hyperspectral Band Reduction


Title	Segmented and Non-Segmented Stacked Denoising Autoencoder for Hyperspectral Band Reduction
Authors	Muhammad Ahmad, Asad Khan, Adil Mehmood Khan, Rasheed Hussain
Abstract	Hyperspectral image analysis often requires selecting the most informative bands instead of processing the whole data without losing the key information. Existing band reduction (BR) methods have the capability to reveal the nonlinear properties exhibited in the data but at the expense of loosing its original representation. To cope with the said issue, an unsupervised non-linear segmented and non-segmented stacked denoising autoencoder (UDAE) based BR method is proposed. Our aim is to find an optimal mapping and construct a lower-dimensional space that has a similar structure to the original data with least reconstruction error. The proposed method first confronts the original hyperspectral data into smaller regions in a spatial domain and then each region is processed by UDAE individually. This results in reduced complexity and improved efficiency of BR for both semi-supervised and unsupervised tasks, i.e. classification and clustering. Our experiments on publicly available hyperspectral datasets with various types of classifiers demonstrate the effectiveness of UDAE method which equates favorably with other state-of-the-art dimensionality reduction and BR methods.
Tasks	Denoising, Dimensionality Reduction
Published	2017-05-19
URL	http://arxiv.org/abs/1705.06920v5
PDF	http://arxiv.org/pdf/1705.06920v5.pdf
PWC	https://paperswithcode.com/paper/segmented-and-non-segmented-stacked-denoising
Repo
Framework

Emergence of Language with Multi-agent Games: Learning to Communicate with Sequences of Symbols


Title	Emergence of Language with Multi-agent Games: Learning to Communicate with Sequences of Symbols
Authors	Serhii Havrylov, Ivan Titov
Abstract	Learning to communicate through interaction, rather than relying on explicit supervision, is often considered a prerequisite for developing a general AI. We study a setting where two agents engage in playing a referential game and, from scratch, develop a communication protocol necessary to succeed in this game. Unlike previous work, we require that messages they exchange, both at train and test time, are in the form of a language (i.e. sequences of discrete symbols). We compare a reinforcement learning approach and one using a differentiable relaxation (straight-through Gumbel-softmax estimator) and observe that the latter is much faster to converge and it results in more effective protocols. Interestingly, we also observe that the protocol we induce by optimizing the communication success exhibits a degree of compositionality and variability (i.e. the same information can be phrased in different ways), both properties characteristic of natural languages. As the ultimate goal is to ensure that communication is accomplished in natural language, we also perform experiments where we inject prior information about natural language into our model and study properties of the resulting protocol.
Tasks
Published	2017-05-31
URL	http://arxiv.org/abs/1705.11192v2
PDF	http://arxiv.org/pdf/1705.11192v2.pdf
PWC	https://paperswithcode.com/paper/emergence-of-language-with-multi-agent-games
Repo
Framework

Design and development of a software system for swarm intelligence based research studies


Title	Design and development of a software system for swarm intelligence based research studies
Authors	Utku Kose
Abstract	This paper introduce a software system including widely-used Swarm Intelligence algorithms or approaches to be used for the related scientific research studies associated with the subject area. The programmatic infrastructure of the system allows working on a fast, easy-to-use, interactive platform to perform Swarm Intelligence based studies in a more effective, efficient and accurate way. In this sense, the system employs all of the necessary controls for the algorithms and it ensures an interactive platform on which computer users can perform studies on a wide spectrum of solution approaches associated with simple and also more advanced problems.
Tasks
Published	2017-04-03
URL	http://arxiv.org/abs/1704.00795v1
PDF	http://arxiv.org/pdf/1704.00795v1.pdf
PWC	https://paperswithcode.com/paper/design-and-development-of-a-software-system
Repo
Framework

Efficient Decision Trees for Multi-class Support Vector Machines Using Entropy and Generalization Error Estimation


Title	Efficient Decision Trees for Multi-class Support Vector Machines Using Entropy and Generalization Error Estimation
Authors	Pittipol Kantavat, Boonserm Kijsirikul, Patoomsiri Songsiri, Ken-ichi Fukui, Masayuki Numao
Abstract	We propose new methods for Support Vector Machines (SVMs) using tree architecture for multi-class classi- fication. In each node of the tree, we select an appropriate binary classifier using entropy and generalization error estimation, then group the examples into positive and negative classes based on the selected classi- fier and train a new classifier for use in the classification phase. The proposed methods can work in time complexity between O(log2N) to O(N) where N is the number of classes. We compared the performance of our proposed methods to the traditional techniques on the UCI machine learning repository using 10-fold cross-validation. The experimental results show that our proposed methods are very useful for the problems that need fast classification time or problems with a large number of classes as the proposed methods run much faster than the traditional techniques but still provide comparable accuracy.
Tasks
Published	2017-08-28
URL	http://arxiv.org/abs/1708.08231v1
PDF	http://arxiv.org/pdf/1708.08231v1.pdf
PWC	https://paperswithcode.com/paper/efficient-decision-trees-for-multi-class
Repo
Framework

Femoral ROIs and Entropy for Texture-based Detection of Osteoarthritis from High-Resolution Knee Radiographs


Title	Femoral ROIs and Entropy for Texture-based Detection of Osteoarthritis from High-Resolution Knee Radiographs
Authors	Jiří Hladůvka, Bui Thi Mai Phuong, Richard Ljuhar, Davul Ljuhar, Ana M Rodrigues, Jaime C Branco, Helena Canhão
Abstract	The relationship between knee osteoarthritis progression and changes in tibial bone structure has long been recognized and various texture descriptors have been proposed to detect early osteoarthritis (OA) from radiographs. This work aims to investigate (1) femoral textures as an OA indicator and (2) the potential of entropy as a computationally efficient alternative to established texture descriptors. We design a robust semi-automatically placed layout for regions of interest (ROI), compute the Hurst coefficient and the entropy in each ROI, and employ statistical and machine learning methods to evaluate feature combinations. Based on 153 high-resolution radiographs, our results identify medial femur as an effective univariate descriptor, with significance comparable to medial tibia. Entropy is shown to contribute to classification performance. A linear five-feature classifier combining femur, entropic and standard texture descriptors, achieves AUC of 0.85, outperforming the state-of-the-art by roughly 0.1.
Tasks
Published	2017-03-27
URL	http://arxiv.org/abs/1703.09296v1
PDF	http://arxiv.org/pdf/1703.09296v1.pdf
PWC	https://paperswithcode.com/paper/femoral-rois-and-entropy-for-texture-based
Repo
Framework

AMR-to-text Generation with Synchronous Node Replacement Grammar


Title	AMR-to-text Generation with Synchronous Node Replacement Grammar
Authors	Linfeng Song, Xiaochang Peng, Yue Zhang, Zhiguo Wang, Daniel Gildea
Abstract	This paper addresses the task of AMR-to-text generation by leveraging synchronous node replacement grammar. During training, graph-to-string rules are learned using a heuristic extraction algorithm. At test time, a graph transducer is applied to collapse input AMRs and generate output sentences. Evaluated on SemEval-2016 Task 8, our method gives a BLEU score of 25.62, which is the best reported so far.
Tasks	Text Generation
Published	2017-02-01
URL	http://arxiv.org/abs/1702.00500v4
PDF	http://arxiv.org/pdf/1702.00500v4.pdf
PWC	https://paperswithcode.com/paper/amr-to-text-generation-with-synchronous-node
Repo
Framework

Statistical Inference for the Population Landscape via Moment Adjusted Stochastic Gradients


Title	Statistical Inference for the Population Landscape via Moment Adjusted Stochastic Gradients
Authors	Tengyuan Liang, Weijie Su
Abstract	Modern statistical inference tasks often require iterative optimization methods to compute the solution. Convergence analysis from an optimization viewpoint only informs us how well the solution is approximated numerically but overlooks the sampling nature of the data. In contrast, recognizing the randomness in the data, statisticians are keen to provide uncertainty quantification, or confidence, for the solution obtained using iterative optimization methods. This paper makes progress along this direction by introducing the moment-adjusted stochastic gradient descents, a new stochastic optimization method for statistical inference. We establish non-asymptotic theory that characterizes the statistical distribution for certain iterative methods with optimization guarantees. On the statistical front, the theory allows for model mis-specification, with very mild conditions on the data. For optimization, the theory is flexible for both convex and non-convex cases. Remarkably, the moment-adjusting idea motivated from “error standardization” in statistics achieves a similar effect as acceleration in first-order optimization methods used to fit generalized linear models. We also demonstrate this acceleration effect in the non-convex setting through numerical experiments.
Tasks	Stochastic Optimization
Published	2017-12-20
URL	http://arxiv.org/abs/1712.07519v2
PDF	http://arxiv.org/pdf/1712.07519v2.pdf
PWC	https://paperswithcode.com/paper/statistical-inference-for-the-population
Repo
Framework

Exploiting Web Images for Weakly Supervised Object Detection


Title	Exploiting Web Images for Weakly Supervised Object Detection
Authors	Qingyi Tao, Hao Yang, Jianfei Cai
Abstract	In recent years, the performance of object detection has advanced significantly with the evolving deep convolutional neural networks. However, the state-of-the-art object detection methods still rely on accurate bounding box annotations that require extensive human labelling. Object detection without bounding box annotations, i.e, weakly supervised detection methods, are still lagging far behind. As weakly supervised detection only uses image level labels and does not require the ground truth of bounding box location and label of each object in an image, it is generally very difficult to distill knowledge of the actual appearances of objects. Inspired by curriculum learning, this paper proposes an easy-to-hard knowledge transfer scheme that incorporates easy web images to provide prior knowledge of object appearance as a good starting point. While exploiting large-scale free web imagery, we introduce a sophisticated labour free method to construct a web dataset with good diversity in object appearance. After that, semantic relevance and distribution relevance are introduced and utilized in the proposed curriculum training scheme. Our end-to-end learning with the constructed web data achieves remarkable improvement across most object classes especially for the classes that are often considered hard in other works.
Tasks	Object Detection, Transfer Learning, Weakly Supervised Object Detection
Published	2017-07-27
URL	http://arxiv.org/abs/1707.08721v2
PDF	http://arxiv.org/pdf/1707.08721v2.pdf
PWC	https://paperswithcode.com/paper/exploiting-web-images-for-weakly-supervised
Repo
Framework