检索结果-内蒙古大学图书馆

IEEE WIC ACM International Conference on Web Intelligence (WI)

作者： Bo Fan Xiaokang Yang Xi Zhou Weiyao Lin Changjian Chen Institute of Image Communication and Network Engineering Shanghai Key Laboratories of Digital Media Processing and Communication Shanghai Jiaotong University Shanghai China Chongqing Institute of Green and Intelligent Technology Chinese Academy of Sciences Chongqing China

In this paper, we proposed an improved coarse to fine improved algorithm to enhance the accuracy of facial key landmark points locating. Based on the analysis of PCA, the proposed algorithm redesigns the parameter update rule through adding a monotonically decreasing inert factor function to the traditional ASM iterations (D-ASM). The new rule could update parameters at a finer process. Besides, we compare the performances of different types of inert factor functions and select the suitable one. Furthermore, we further design a classifier-based algorithm for the more accurate locating of 2D key corner points. Finally, local D-ASM is constructed and the inner landmarks are further fitting with corner points fixed. Experimental results on various faces demonstrate the effectiveness and rationality of our proposed algorithm.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cooperative MIMO Channel Modeling and Multi-Link Spatial Correlation Properties

引用

IEEE Journal on Selected Areas in Communications 2012年第2期30卷 388-396页

作者： Xiang Cheng Cheng-Xiang Wang Haiming Wang Xiqi Gao Xiao-Hu You Dongfeng Yuan Bo Ai Qiang Huo Ling-Yang Song Bing-Li Jiao School of Electronics and Computer Science Peking University Beijing China Key Laboratory of Wireless Sensor Network & Communication Shanghai Institute of Microsystem and Information Technology Chinese Academy of Sciences China State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University China Joint Research Institute for Signal and Image Processing School of Engineering & Physical Sciences Heriot-Watt University Edinburgh UK School of Information Science and Engineering South East University Najing China School of Information Science and Engineering Shandong University Jinan China State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University Beijing China

In this paper, a novel unified channel model framework is proposed for cooperative multiple-input multiple-output (MIMO) wireless channels. The proposed model framework is generic and adaptable to multiple cooperative MIMO scenarios by simply adjusting key model parameters. Based on the proposed model framework and using a typical cooperative MIMO communication environment as an example, we derive a novel geometry-based stochastic model (GBSM) applicable to multiple wireless propagation scenarios. The proposed GBSM is the first cooperative MIMO channel model that has the ability to investigate the impact of the local scattering density (LSD) on channel characteristics. From the derived GBSM, the corresponding multi-link spatial correlation functions are derived and numerically analyzed in detail.

关键词： MIMO Correlation Channel models Scattering Educational institutions Fading Adaptation models

来源：评论

学校读者我要写书评

暂无评论

A survey of medical applications of 3D image analysis and computer graphics

引用

Systems and Computers in Japan 2006年第1期37卷 13-46页

作者： Muraki, Shigeru Kita, Yasuyo National Institute of Advanced Industrial Science and Technology Tokyo 135-0064 Japan Information Technology Research Institute National Institute of Advanced Industrial Science and Technology Tsukuba 305-8568 Japan Video Information Media Society Japan Neural Network Society ACM IEEE Computer Society Information Technology Research Institute National Institute of Advanced Industrial Science and Technology Information Processing Society Japan Medical Image Engineering Society

This paper is a survey of visualization and analysis techniques for medical 3D images for researchers and students in computer science. Publications from this decade with internationally high evaluation are reviewed, focusing on medical applications of new mathematical techniques and computer hardware, mostly from the viewpoint of the authors' specialty, namely, volume graphics and computer vision. © 2005 Wiley Periodicals, Inc.

关键词： image analysis

来源：评论

学校读者我要写书评

暂无评论

Analysis of Japanese dance movements using motion capture system

引用

Systems and Computers in Japan 2006年第1期37卷 71-82页

作者： Yoshimura, Mitsu Murasato, Hideki Kai, Tamiko Kuromiya, Akira Yokoyama, Kiyoko Hachimura, Kozaburo Center for Promotion of the COE Ritsumeikan University Kyoto 603-8577 Japan Graduate School of Design and Architecture Nagoya City University Nagoya 464-0083 Japan Nagoya Municipal Industrial Research Institute Nagoya 456-0058 Japan College of Science and Engineering Ritsumeikan University Kusatsu 525-8577 Japan Information Processing Society of Japan ATR Network Information Laboratories Graduate School of Design and Architecture Nagoya City University Institute of Electrical Engineers of Japan Japan Ergonomics Society Japan Society for Medical and Biological Engineering Department of Informatics College of Science and Engineering Ritsumeikan University Institute of Image Electronics Engineers of Japan

In this research, the authors evaluate the degree to which dancers copy or follow the techniques of a master, or the degree of proficiency, by analyzing movements in traditional Japanese dance. The data used consist of three-dimensional time series for traditional Japanese dance movements acquired using optical motion capture system. In the authors' prior research, three moving coordinate systems which would move according to the translation and rotation of the body were used to extract the portion of the target movement. In this research, the authors consider a moving coordinate system which simultaneously takes into consideration translation, rotation, correction of orientation, and correction of waist tremble. In their prior research, the authors defined indices for movement stability and frequency characteristics as indices to quantitatively represent in an objective fashion the degree of proficiency of a dancer. Separate from this, in the current research the authors define an index with the spectrum component using a Gabor transform and an index for the amount of translation. The authors had a total of five people, a master from a particular dance school and four dance students of different genders and at different experience levels (all the master's students), perform dance experiments. The authors then extracted the target movements, measured the indices using the extraction results, and attempted to evaluate the degree of proficiency based on the proposed indices. Extraction was sufficiently precise, and the authors were able to confirm that the indices represent the differences appearing due to degree of proficiency and gender. © 2005 Wiley Periodicals, Inc.

关键词： Motion estimation

来源：评论

学校读者我要写书评

暂无评论

A detection method of moving ships by image processing as a support system of AIS

A detection method of moving ships by image processing as a ...

引用

institute of Navigation, 2005 National Technical Meeting, NTM 2005

作者： Shimpo, M. Hirasawa, M. Nakajima, A. Shoji, R. Oshima, M. Tokyo University of Marine Science and Technology Information Systems Planning and Development Department Mitsubishi Electric Corporation Japan Institute of Navigation Robotics Society of Japan Institute of Electronics Information and Communication Engineers Japan Institute of Navigation Image Processing Radar Network System Faculty of Marine Technology Tokyo University of Marine Science and Technology Electro Technical Laboratory Ibaraki Japan Tokyo University of Mercantile Marine Tokyo Japan Institute of Electronics Information and Communication Engineers Information Processing Society of Japan Japanese Society for Artificial Intelligence Marine Engineering Society in Japan Law and Computers Association of Japan

This paper proposes a detection method of moving ships from the navigational image sequence that was taken with cameras installed on the bridge of the ship. The image is influenced by roll and pitch of the ship. Therefore, usual technique of image processing cannot be used such as Finite Difference method. Moreover, as many sea waves appear in the images, the image cannot be dealt without rejecting sea waves. The technique in this paper proposes how to get rid of the influence of roll, pitch and sea waves. After that, we detect the ships. An interval between each of the frames is one-thirty second in this experiment. The size of the images is 640 pixels in width and 480 pixels in height. At first, the frames are segmented into about 5000 regions using brightness value of the image. About 100 regions were estimated with the ship. Each region was matched with another two frames, which passed 0.33 second and 0.66 second. 0.33 second corresponds to 10 frames. The movement of the ship is less than 50 pixels in this interval. The SSDA method is used for matching processing to be more efficiently. Both the coordinate points of an original region and the matched one are recorded on the table. It is used to calculate the deflection, the speed and the direction of the moving vector of each region. The ships can be detected with these parameters. The detected ships can be displayed clearly as a result of processing. We have taken a lot of video images in Tokyo Bay and Tokyo Port. It becomes possible to show that the ships can be detected from the video images.

关键词： Navigation systems

来源：评论

学校读者我要写书评

暂无评论

Efficient Approach for Face Detection in Video Surveillance

引用

Journal of Donghua University(English Edition) 2003年第4期20卷 52-55页

作者：宋红石峰 Department of Computer Science and Engineering Beijing Institute of Technology Beijing 100081 China Department of Computer Science and Engineering Beijing Institute of Technology Beijing 100081 Chinaecurity access control systems and automatic video surveillance systems are becoming increasingly important recently and detecting human faces is one of the indispensable processes. In this paper an approach is presented to detect faces in video surveillance. Firstly both the skin-color and motion components are applied to extract skin-like regions. The skin-color segmentation algorithm is based on the BPNN (back-error-propagation neural network) and the motion component is obtained with frame difference algorithm. Secondly the image is clustered into separated face candidates by using the region growing technique. Finally the face candidates are further verified by the rule-based algorithm. Experiment results demonstrate that both the accuracy and processing speed are very promising and the approach can be applied for the practical use.

Security access control systems and automatic video surveillance systems are becoming increasingly important recently,and detecting human faces is one of the indispensable *** this paper,an approach is presented to detect faces in video ***,both the skin-color and motion components are applied to extract skin-like *** skin-color segmentation algorithm is based on the BPNN (back-error-propagation neural network) and the motion component is obtained with frame difference ***,the image is clustered into separated face candidates by using the region growing ***,the face candidates are further verified by the rule-based *** results demonstrate that both the accuracy and processing speed are very promising and the approach can be applied for the practical use.

关键词： face detection skin-color segmentation BPNN frame difference region growing

来源：评论

学校读者我要写书评

暂无评论

Using sinusoidal spectral power distributions to evaluate color scanning filters

引用

Journal of the Society for Information Display 1998年第4期6卷 299-305页

作者： Kotera, Hiroaki Fumoto, Teruo Yoshida, Kunio Dept. of Info. and Image Sciences Chiba University 1-33 Yayoicho Inage-ku Chiba 263 Japan Matsushita Res. Inst. Tokyo Inc. 3-10-1 Higashimita Tama-ku Kawasaki 214 Japan Nagoya Institute of Technology University of Tokyo Matsushita Elec. Indust. Co. Ltd. Research Institute Tokyo Inc. Chiba University Dept. of Info. and Image Science Faculty of Engineering Osaka University Matsushita Res. Inst. Tokyo Inc. Inst. Television Engineers of Japan Inst. Image Electron. Engineers J. Waseda University International Neural Network Society Info. Processing Society of Japan

We demonstrate a method for evaluating the quality of color scanning-filter sets that use sinusoidal spectral power distributions (sine SPDs) instead of physical test targets. Filter quality is quantified as the mean square error of the filter sets' responses to the fundamental metamers of sine SPDs having varying frequency and phase, relative to a perfect filter set. Filter quality is also depicted graphically by plotting filter input versus output in CIELAB color space, and by plotting the magnitude of the filters' CIELAB color vector response to sine SPDs. The advantages of this approach to scanning-filter evaluation are discussed.

关键词： Optical filters

来源：评论

学校读者我要写书评

暂无评论

ReAL: Improving image-Text Retrieval with Authentic Negative Repository Learning

引用

ACM Transactions on Multimedia Computing, Communications, and Applications 1000年

作者： Renjie Pan Hua Yang Xiangyu Zhao Institute of Image Communication and Network Engineering Shanghai Key Lab of Digital Media Processing and Transmission Shanghai Jiao Tong University China Institute of Image Communication and Network Engineering Shanghai Jiao Tong University China

Current methods for image-text retrieval commonly propose various fusion modules to achieve robust visual-textual alignment, primarily relying on in-batch learning to guide the matching process. Some follow-up methods seek to enlarge the number of negative samples to boost image-text contrastive learning. However, these methods often face challenges posed by semantic-consistent negatives, i.e., negatives samples that share correspondence with the ground truth, leading to confusion in learning cross-modal semantics. To address this issue, we propose a novel Retrieve with Authentic negative repository Learning (ReAL) method, which constructs a specific Authentic Negative Repository filled with valuable negative sample pairs. By introducing a Unique Negative Filter with a Discriminative Triplet Ranking Loss, ReAL effectively filters out the semantic-consistent negatives through similarity distribution analysis and threshold learning. Moreover, existing fusion paradigms suffer from intricate use of fine-grained representations from word- and region-level instances to progressively refine the fused embedding. In this paper, we propose a lightweight Cluster Refinement Module to exploit cross-modal semantics in a 1-way-1-out paradigm. Each visual-textual alignment can spontaneously uncover correlations with adjacent alignments through aggregation and re-allocation, without the need for a redundant and cost-inefficient refinement stage. Furthermore, ReAL employs dual momentum encoders with two memory banks, expanding the selection range of the Authentic Negative Repository to include a broader set of negatives. Extensive experiments conducted on Flickr30K, MS-COCO, and the augmented Flickr30K (with more hard negatives) demonstrate the superiority and robustness of ReAL, while also showcasing its significantly reduced inference time compared to other competitive baselines.

关键词： image-text Retrieval Authentic Negative Repository Cross-modal Fusion

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：