检索结果-内蒙古大学图书馆

7th International conference on Internet Multimedia Computing and Service, ICIMCS 2015

作者： Fu, Wei Zhao, Chaoyang Wang, Jinqiao Liu, Jing Cheng, Jian Lu, Hanqing 54th Institute of China Electronics Technology Group Corporation Shijiazhuang Hebei050081 China National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing100190 China

ISBN: (纸本)9781450335287

Group activity classification is the task to identify activities with multiple person participation, which often involves in the usage of the context information like person relation- ships and person interactions. In this paper, we propose a novel approach to jointly model three co-existing cues in- cluding the activity duration time, individual action feature and the context information shared between person interac- tions. Our approach infers group activity labels of all the persons together with their activity durations, especially for the situation with multiple group activities co-existing. Ex- perimental results show that our approach outperform state- of-the-art by 10%. © 2015 ACM.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Preface

Communications in Computer and Information Science

引用

Communications in computer and Information Science 2016年 662卷

作者： Tan, Tieniu Li, Xuelong Chen, Xilin Zhou, Jie Yang, Jian Cheng, Hong Chinese Academy of Sciences Institute of Automation Beijing China Xi’an Institute of Optics and Precision Mechanics Chinese Academy of Sciences Xi’an China Chinese Academy of Sciences Institute of Computing Technology Beijing China Tsinghua University Beijing China Nanjing University of Science and Technology Nanjing China University of Electronic Science University of Electronic Science ChengduSichuan China

来源：评论

学校读者我要写书评

暂无评论

PE-TLD: Parallel extended tracking-learning-detection for multi-target tracking 15th

PE-TLD: Parallel extended tracking-learning-detection for mu...

引用

15th International conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2015

作者： Zhou, Chenggang Dong, Qiankun Ma, Wenjing Long, Guoping Li, Tao College of Computer and Control Engineering Nankai University Tianjin300350 China Institute of Software Chinese Academy of Sciences Beijing100190 China

ISBN: (纸本)9783319271217

Multi-target tracking in video has been a research focus, with the combination of many fields, such as computer vision, artificial intelligence, pattern matching. In this paper, we present an efficient multi-target recognition and tracking algorithm based on TLD (Tracking-Learning-Detection), named PETLD. A new foreground extraction filter using ViBe is introduced to improve the speed and accuracy of detection. A new target recognition component is added, and core detector is improved. Based on that, we further implemented a parallel version, taking advantage of the state-of-the-art parallel computing techniques such as OpenMP and OpenCL, which runs efficiently on a system with both multicore CPU and GPU. Experiments showed that PE-TLD is up to 5 times faster than the serial version. PE-TLD is an automatic multi-target recognition and tracking system, which is efficient enough to be deployed for real-time usage. © Springer International Publishing Switzerland 2015.

关键词： pattern matching

来源：评论

学校读者我要写书评

暂无评论

Adaptive Structural Model for Video Based Pedestrian Detection 12

引用

12th Asian conference on computer vision (ACCV)

作者： Yan, Junjie Yang, Bin Lei, Zhen Li, Stan Z. Chinese Acad Sci Ctr Biometr & Secur Res Beijing Peoples R China Chinese Acad Sci Natl Lab Pattern Recognit Ctr Biometr & Secur Res Beijing Peoples R China

ISBN: (纸本)9783319168654;9783319168647

the performance of generic pedestrian detector usually declines seriously for videos in novel scenes, which is one of the major bottlenecks for current pedestrian detection techniques. the conventional works improve pedestrian detection in video by mining new instances from detections and adapting the detector according to the collected instances. However, when treating the two tasks separately, the detector adaptation suffers from the defective output of instance mining. In this paper, we propose to jointly handle the instance mining and detector adaption using an adaptive structural model. the regularization function of the model is applied on detector to prevent overfitting in adaption, and the loss function is designed to evaluate the combination of mined instances set and detector. Particularly, we extend the Deformable Part Model (DPM) to adaptive DPM, where an adaptive feature transformation defined on low-level HOG cell is learned to reduce the domain shift, and the regularization function for the detector is conducted on the transformation. the loss of the instance set and detector is measured by a cost-flow network structure which incorporates both the appearance of frame-wise detections and their spatio-temporal continuity. We demonstrate an alternating minimization procedure to optimize the model. the proposed method is evaluated on EthZ, PETS2009 and Caltech datasets, and outperforms baseline DPM by 7% in terms of mean miss rate.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Fast Convolution Operations on Many-Core Architectures 17

Fast Convolution Operations on Many-Core Architectures

引用

2015 IEEE 17th International conference on High Performance Computing and Communications (HPCC)

作者： Li, Shigang Zhang, Yunquan Xiang, Chunyang Shi, Lei Chinese Acad Sci Inst Comp Technol State Key Lab Comp Architecture Beijing Peoples R China Zhengzhou Univ Sch Informat Engn Zhengzhou Peoples R China

ISBN: (纸本)9781479989379

Convolution operations have been widely used in many important application domains, such as deep learning and computer vision, in which convolution is always the most time-consuming part. High computational throughput and memory bandwidth make many-core architectures the promising targets to accelerate these applications. In this paper, we implement and optimize different convolution operations, including 1D convolution, 2D convolution and multi-channel 2D convolution executed in mini-batch mode, on both GPU and Intel MIC many-core architectures. We find out that the performance bottleneck of 1D and 2D convolutions is on registers rather than local memory or L1/L2 cache, and therefore, register tiling is used to improve the performance. In addition, we present a novel solution for multi-channel 2D convolution, in which convolution is conducted on images directly instead of being translated to matrix multiplication, and the data reuse of the algorithm is fully exploited. We further summarize the parameters of autotuning for multi-channel 2D convolution and prune the search space based on heuristics. the experimental results show that, for the large filter size, our solution gets up to 33% performance improvement over cuDNN-v2 and up to 28% over clBLAS-based implementation, on GTX TITAN and AMD W8000 respectively. On Intel MIC, our solution gets up to 25% of the theoretical peak performance.

关键词： Convolution GPU Intel MIC OpenCL Deep learning computer vision

来源：评论

学校读者我要写书评

暂无评论

Preliminary study on QR code detection using HOG and AdaBoost

Preliminary study on QR code detection using HOG and AdaBoos...

引用

International conference of Soft Computing and pattern recognition (SOCPAR)

作者： Yih-Lon Lin Chung-Ming Sung Department of Information Engineering I-Shou University Kaohsiung Taiwan

ISBN: (纸本)9781467393614

In this paper, an approach of QR code detection using histograms of oriented gradients (HOG) and AdaBoost is proposed. there are two steps in our approach. In the first step, feature vectors are extracted using HOG with various cell sizes and overlapping or non-overlapping blocks. In the second step, the AdaBoost algorithms are trained by the input feature vectors from HOG and output targets. the QR code position is then detected via the predicted outputs from the AdaBoost algorithm. Experimental results show that the proposed method is an effective way to detect QR code position. Frankly speaking, the results reported here only provide preliminary study on QR code detection using HOG and AdaBoost.

关键词： Feature extraction Histograms Training Image edge detection computer vision Object detection pattern recognition

来源：评论

学校读者我要写书评

暂无评论

7th International conference on Social Informatics, SocInfo 2015

7th International Conference on Social Informatics, SocInfo ...

引用

7th International conference on Social Informatics, SocInfo 2015

ISBN: (纸本)9783319274324

the proceedings contain 19 papers. the special focus in this conference is on Social Informatics. the topics include: Culture, imagined audience, and language choices of multilingual chinese and korean students on facebook;analyzing factors impacting revining on the vine social network;analyzing labeled cyberbullying incidents on the instagram social network;uncovering social media reaction pattern to protest events;detecting opinions in a temporally evolving conversation on twitter;identifying similar opinions in news comments using a community detection algorithm;identifying suggestions for improvement of product features from online product reviews;crowdsourcing safety perceptions of people;a real-time crowd-powered testbed for content assessment of potential social media posts;adaptive survey design using structural characteristics of the social network;linking profiles across social networks;choosing the right home location definition method for the given dataset;proposing ties in a dense hypergraph of academics;comparing hypotheses about urban photo trails on flickr;labor saving and labor making of value in online congratulatory messages;banzhaf index for influence maximization and modeling social media content with word vectors for recommendation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

the Impact of Natural Utilization of Traditional chinese Cultural Elements on the User Experience in Mobile Interaction Design 17th

引用

7th International conference on Cross-Cultural Design (CCD) Held as Part of 17th International conference on Human-computer Interaction (HCI International)

作者： Lei, Tian Liu, Xu Wu, Lei Chen, Tianjian Wang, Yuhui Xiong, Luyao Wei, Shuaili Huazhong Univ Sci & Technol Dept Ind Design Wuhan 430074 Peoples R China Carnegie Mellon Univ Dept Mech Engn Pittsburgh PA 15213 USA

ISBN: (纸本)9783319209340;9783319209333

this paper, by making two experiments, studies the impact of natural culture integration on the user experiences. the results of experiment I indicate that: (1) Cloud pattern and Ink Painting are the most likely to activate "sliding" gesture, followed by "tapping" gesture;(2) chinese Seals and Paper-cut tend to activate "pinching" gesture, followed by "sliding" and "tapping" gestures;(3) Gu Zheng and Shadow Figures have generally same tendency to activate "tapping", "sliding" and "pinching" gestures;(4) Calligraphy tends to activate "sliding" gesture. the results of experiment II show that: (1) gesture type has very limited influence on the interaction experience;(2) task complexity affects the user experience significantly;(3) the way of integration of cultural elements has significant relationship with "Feedback Clarity", "Easy-to-use", "Satisfaction in Feedback", "Satisfaction for Culture Utilization", "Natural Degree of Culture Utilization" and "Degree of Interesting", and has very limited correlations with "Feedback Understandability" and "Memorability". Except for "Feedback Clarity", other 5 evaluations are the best in the natural integration condition and the worst in no integration condition. In brief, the two experiments indicate that the natural integration of cultural elements promoted the user experience, even though it reduced the clarity.

关键词： Culture representation User experience Mobile usability

来源：评论

学校读者我要写书评

暂无评论

pattern Characterization and Connectivity Analysis for Edge Feature Extraction 7

Pattern Characterization and Connectivity Analysis for Edge ...

引用

7th International conference on Digital Image Processing (ICDIP)

作者： Jiang, Bo Chinese Acad Sci Guangzhou Inst Biomed & Hlth Div Appl Stem Cell Automat Guangzhou 510530 Guangdong Peoples R China

ISBN: (纸本)9781628418293

the purpose of detecting sharp changes in image brightness is to capture important events and changes in properties of the world. the accuracy of edge detection methods in image processing determines the eventual success or failure of computerized analysis procedures which follow the initial edge detection determinations such as object recognition. Generally, edge detectors have been designed to capture simple ideal step functions in image data, but real image signal discontinuities deviate from this ideal form. Another three types of deviations from the step function which relate to real distortions occurring in natural images are examined according to their characteristics. these types are impulse, ramp, and sigmoid functions which respectively represent narrow line signals, simplified blur effects, and more accurate blur modeling. General rules for edge pattern characterization based upon the classification of edge types into four categoriesramp, impulse, step, and sigmoid (RISS) are developed from this analysis. Additionally, the proposed algorithm performs connectivity analysis on edge map to ensure that small, disconnected edges are removed. the performance analysis on experiments supports that the proposed edge detection algorithm with edge pattern analysis and characterization does lead to more effective edge detection and localization with improved accuracies. To expand the proposed algorithm into real-time applications, a parallel implementation on a graphics processing unit (GPU) is presented in this paper. For the various configurations in our test, the GPU implementation shows a scalable speedup as the resolution of an image increases. We also achieved 14 frames per second in real-time processing (1280x720).

关键词： Feature extraction edge feature edge detection pattern characterization connectivity analysis GPU

来源：评论

学校读者我要写书评

暂无评论

Uncovering Social Media Reaction pattern to Protest Events: A Spatiotemporal Dynamics Perspective of Ferguson Unrest 1

引用

7th International conference on Social Informatics (SocInfo)

作者： He, Jiaying Hong, Lingzi Frias-Martinez, Vanessa Torrens, Paul Univ Maryland Dept Geog Sci College Pk MD 20742 USA Univ Maryland Sch Informat Studies College Pk MD 20742 USA

ISBN: (数字)9783319274331

ISBN: (纸本)9783319274331;9783319274324

Social platforms like Twitter play an important role in people's participation in social events. Utilizing big social media data to uncover people's reaction to social protests can shed lights on understanding the event progress and the attitudes of normal people. In this study, we aim to explore the use of Twitter during protests using Ferguson unrest as an example from multiple perspectives of space, time and content. We conduct an in-depth analysis to unpack the social media response and event dynamics from a spatiotemporal perspective and to evaluate the social media reaction through the integration of spatiotemporal tweeting behavior and tweet text. We propose to answer the following research questions. (1) What is the general spatiotemporal tweeting patterns across the US? (2) What is the spatiotemporal tweeting patterns in local St. Louis? (3) What are the reaction patterns in different US urban areas in space, time and content?

关键词： Social networking (online)

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：