检索结果-内蒙古大学图书馆

Learning a saliency evaluation metric using crowdsourced perceptual judgments

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Xia, Changqun Li, Jia Su, Jinming Borji, Ali State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing China International Research Institute for Multidisciplinary Science Beihang University Beijing China Center for Research in Computer Vision Computer Science Department University of Central Florida OrlandoFL United States

In the area of human fixation prediction, dozens of computational saliency models are proposed to reveal certain saliency characteristics under different assumptions and definitions. As a result, saliency model benchmarking often requires several evaluation metrics to simultaneously assess saliency models from multiple perspectives. However, most computational metrics are not designed to directly measure the perceptual similarity of saliency maps so that the evaluation results may be sometimes inconsistent with the subjective impression. To address this problem, this paper first conducts extensive subjective tests to find out how the visual similarities between saliency maps are perceived by humans. Based on the crowdsourced data collected in these tests, we conclude several key factors in assessing saliency maps and quantize the performance of existing metrics. Inspired by these factors, we propose to learn a saliency evaluation metric based on a two-stream convolutional neural network using crowdsourced perceptual judgements. Specifically, the relative saliency score of each pair from the crowdsourced data is utilized to regularize the network during the training process. By capturing the key factors shared by various subjects in comparing saliency maps, the learned metric better aligns with human perception of saliency maps, making it a good complement to the existing metrics. Experimental results validate that the learned metric can be generalized to the comparisons of saliency maps from new images, new datasets, new models and synthetic data. Due to the effectiveness of the learned metric, it also can be used to facilitate the development of new models for fixation prediction. Copyright © 2018, The Authors. All rights reserved.

关键词： Crowdsourcing

Direction-aware spatial context features for shadow detection and removal

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Hu, Xiaowei Fu, Chi-Wing Zhu, Lei Qin, Jing Heng, Pheng-Ann Department of Computer Science and Engineering Chinese University of Hong Kong Centre for Smart Health School of Nursing Hong Kong Polytechnic University Department of Computer Science and Engineering Chinese University of Hong Kong Guangdong Provincial Key Laboratory of Computer Vision and Virtual Reality Technology Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China

Shadow detection and shadow removal are fundamental and challenging tasks, requiring an understanding of the global image semantics. This paper presents a novel deep neural network design for shadow detection and removal by analyzing the image context in a direction-aware manner. To achieve this, we first formulate the direction-aware attention mechanism in a spatial recurrent neural network (RNN) by introducing attention weights when aggregating spatial context features in the RNN. By learning these weights through training, we can recover direction-aware spatial context (DSC) for detecting and removing shadows. This design is developed into the DSC module and embedded in a convolutional neural network (CNN) to learn the DSC features in different levels. Moreover, we design a weighted cross entropy loss to make effective the training for shadow detection and further adopt the network for shadow removal by using a Euclidean loss function and formulating a color transfer function to address the color and luminosity inconsistency in the training pairs. We employ two shadow detection benchmark datasets and two shadow removal benchmark datasets, and perform various experiments to evaluate our method. Experimental results show that our method clearly outperforms state-of-the-art methods for both shadow detection and shadow removal. Copyright © 2018, The Authors. All rights reserved.

关键词： Deep neural networks

BlindReader: An assistant system to support text reading for the visually impaired people

学校读者我要写书评

暂无评论

Journal of Information Hiding and Multimedia Signal Processing 2018年第4期9卷 938-948页

作者： Guo, Fan He, Han-Dong Xiao, Yan-Ru Fu, Teng-Mu Liao, Cheng Tan, Ping School of Information Science and Engineering Central South University Changsha410083 China Key Laboratory of Hunan Province for New Retail Virtual Reality Technology School of Computer and Information Engineering Hunan University of Commerce Changsha410205 China

Accessing printed text is a great challenge for the visually impaired people. Therefore, in this paper, we introduce a visual assistant system, BlindReader, that assists visually impaired people to read printed text. Some computer vision algorithms are introduced to adjust the intensity and orientation of captured document images in preprocessing stage. Using the system, the low-vision users can read the text more clearly just like the way people use the magnifying glass and the blind users can hear the recognized words synthesized to audio speech for the whole text or the single text line. Finally, the system is evaluated in terms of text recognition accuracy and user feedback to determine the usability of the BlindReader. © 2018, Ubiquitous International. All rights reserved.

关键词： Character recognition

Data-Driven Hair Modeling from a Single Image

学校读者我要写书评

暂无评论

Data-Driven Hair Modeling from a Single Image

International Conference on virtual reality and Visualization (ICVRV)

作者： Jiqiang Wu Yongtang Bao Yue Qi State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing China College of Computer Science and Engineering Shandong University of Science and Technology Qingdao China Beihang University Qingdao Research Institute Qingdao China

ISBN: (纸本)9781538684986;9781538684979

Hair is one of the most distinctive part of a human body, which is essential for the digitization of compelling virtual avatars. We present a data-driven approach for generating complete and complex 3D hairstyles from a single-view portrait. We construct a hairstyle database which contains more than 2000 hair models. Given a target hairstyle portrait image, we first draw a few strokes as guidance. We then search multiple best matching examples from the database based on our enhanced matching algorithm. Finally, we combine them consistently into a single hairstyle. The generated hairstyles are visually comparable to original portrait images. The reconstructed 3D hair models can be used for many applications, such as hair editing, and dynamic hair simulation.

关键词： Hair Solid modeling Three-dimensional displays Databases Computational modeling Two dimensional displays Head

Anchor-based nearest class mean loss for convolutional neural networks

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Hao, Fusheng Cheng, Jun Wang, Lei Wang, Xinchao Cao, Jianzhong Hu, Xiping Tao, Dapeng Xi'an Institute of Optics and Precision Mechanics CAS China University of Chinese Academy of Sciences Shenzhen Key Laboratory of Virtual Reality and Human Interaction Technology Shenzhen Institutes of Advanced Technology CAS Department of Computer Science Stevens Insititute of Technology

Discriminative features are critical for machine learning applications. Most existing deep learning approaches, however, rely on convolutional neural networks (CNNs) for learning features, whose discriminant power is not explicitly enforced. In this paper, we propose a novel approach to train deep CNNs by imposing the intra-class compactness and the inter-class separability, so as to enhance the learned features' discriminant power. To this end, we introduce anchors, which are predefined vectors regarded as the centers for each class and fixed during training. Discriminative features are obtained by constraining the deep CNNs to map training samples to the corresponding anchors as close as possible. We propose two principles to select the anchors, and measure the proximity of two points using the Euclidean and cosine distance metric functions, which results in two novel loss functions. These loss functions require no sample pairs or triplets and can be efficiently optimized by batch stochastic gradient descent. We test the proposed method on three benchmark image classification datasets and demonstrate its promising results. Copyright © 2018, The Authors. All rights reserved.

关键词： Anchors

Attentive relational networks for mapping images to scene graphs

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Qi, Mengshi Li, Weijian Yang, Zhengyuan Wang, Yunhong Luo, Jiebo State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University China Beijing Advanced Innovation Center for Big Data and Brain Computing Department of Computer Science University of Rochester United States

Scene graph generation refers to the task of automatically mapping an image into a semantic structural graph, which requires correctly labeling each extracted object and their interaction relationships. Despite the recent success in object detection using deep learning techniques, inferring complex contextual relationships and structured graph representations from visual data remains a challenging topic. In this study, we propose a novel Attentive Relational Network that consists of two key modules with an object detection backbone to approach this problem. The first module is a semantic transformation module utilized to capture semantic embedded relation features, by translating visual features and linguistic features into a common semantic space. The other module is a graph self-attention module introduced to embed a joint graph representation through assigning various importance weights to neighboring nodes. Finally, accurate scene graphs are produced by the relation inference module to recognize all entities and the corresponding relations. We evaluate our proposed method on the widely-adopted Visual Genome Dataset, and the results demonstrate the effectiveness and superiority of our model. Copyright © 2018, The Authors. All rights reserved.

关键词： Semantics

Computing restricted Voronoi diagram on graphics hardware 25

学校读者我要写书评

暂无评论

Computing restricted Voronoi diagram on graphics hardware

25th Pacific Conference on computer Graphics and Applications, Pacific Graphics 2017

作者： Han, Jiawei Yan, Dong-Ming Wang, Lili Zhao, Qinping State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University China National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences China

ISBN: (纸本)9783038680512

The 3D restricted Voronoi diagram (RVD), defined as the intersection of the 3D Voronoi diagram of a pointset with a mesh surface, has many applications in geometry processing. There exist several CPU algorithms for computing RVDs. However, such algorithms still cannot compute RVDs in realtime. In this short paper, we propose an efficient algorithm for computing RVDs on graphics hardware. We demonstrate the robustness and the efficiency of the proposed GPU algorithm by applying it to surface remeshing based on centroidal Voronoi tessellation. © 2017 The Author(s) Eurographics Proceedings © 2017 The Eurographics Association.

关键词： Computational geometry

Image co-segmentation via multi-scale local shape transfer

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Teng, Wei Zhang, Yu Chen, Xiaowu Li, Jia He, Zhiqiang State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China International Research Institute for Multidisciplinary Science Beihang University Beijing100191 China Lenovo Research China

Image co-segmentation is a challenging task in computer vision that aims to segment all pixels of the objects from a predefined semantic category. In real-world cases, however, common foreground objects often vary greatly in appearance, making their global shapes highly inconsistent across images and difficult to be segmented. To address this problem, this paper proposes a novel co-segmentation approach that transfers patch-level local object shapes which appear more consistent across different images. In our framework, a multi-scale patch neighbourhood system is first generated using proposal flow on arbitrary image-pair, which is further refined by Locally Linear Embedding. Based on the patch relationships, we propose an efficient algorithm to jointly segment the objects in each image while transferring their local shapes across different images. Extensive experiments demonstrate that the proposed method can robustly and effectively segment common objects from an image set. On iCoseg, MSRC and Coseg-Rep dataset, the proposed approach performs comparable or better than the state-of-the-arts, while on a more challenging benchmark Fashionista dataset, our method achieves significant improvements. Copyright © 2018, The Authors. All rights reserved.

关键词： Semantics

Moving object detection algorithm based on pixel background sample sets in panoramic scanning mode 17

学校读者我要写书评

暂无评论

Moving object detection algorithm based on pixel background ...

2017 International Conference on Compute and Data Analysis, ICCDA 2017

作者： Zhang, Chi Zheng, Jin Zhang, Yugui Zhang, Zhi Beijing Key Laboratory of Digital Media School of Computer Science and Engineering Beihang University Beijing100191 China Beijing Key Laboratory of Digital Media State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China

ISBN: (纸本)9781450352413

In order to overcome the excessive false detection of marginal noise and the object holes of the existing algorithm in outdoor panoramic surveillance, a moving object detection algorithm based on pixel background sample sets in panoramic scanning mode is proposed. In the light of the space distribution characteristics, neighborhood pixels have similar values. Therefore, a background sample set for each pixel is created by random sampling in the first scanning cycle which effectively avoids the false detection of marginal noise and reduces the time cost of background model establishment. The adjacent frame difference detection algorithm in the traditional camera motion mode is prone to object holes. To solve this problem, detection based on background sample sets is presented to obtain complete moving object region. The results indicate that the proposed moving object detection algorithm works more efficiently on reducing marginal noise interference, and obtains complete moving object information compared with the frame difference detection algorithm based on registration results in traditional camera motion mode, thereby meeting the needs of real-time detection as well as improving its accuracy. Copyright 2017 ACM.

关键词： Pixels