检索结果-内蒙古大学图书馆

3rd Asian Conference on Computer Vision, ACCV 1998

作者： Tong, W.B. Tso, S.K. Lang, S. Lu, G.Z. Ma, S.D. Centre for Intelligent Design Automation and Manufacturing City University of Hong Kong Tat Chee Avenue Kowloon Hong Kong AI and Robotics Lab Dept. Of Computer and System Science Nankai University Tianjing300071 China National Lab of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing100080 China

ISBN: (纸本)3540639306

For most present-day industrial robots, the user programs the robot using a robot language or teaches the robot using a pendant and/or joystick. In telerobots, the user controls the mobile robot or the robot arm by manually operating a joystick where the user himself acts as a cognition and decision making unit in the control loop. One of the important sensed information given to the user is visual feedback. Inspired by behavioral control of mobile robots, we propose a novel interface between the user and the robot to improve the performance of the robot system. In the proposed paradigm, the visual feedback exists in two control loops: one for the human supervisory control and the other for the robot itself for visual servoing. The human operator only provides the visual primitives to the robot Bad tells the robot its current situation which will activate a certain action, as in situation-to-action reactive/reflex control of mobile robots. © Springer-Verlag Berlin Heidelberg 1997.

关键词： Mobile robots

来源：评论

学校读者我要写书评

暂无评论

FICAL: Focal Inter-Class Angular Loss for Image Classification

FICAL: Focal Inter-Class Angular Loss for Image Classificati...

引用

IEEE Visual Communications and Image Processing Conference

作者： Xinran Wei Dongliang Chang Jiyang Xie Yixiao Zheng Chen Gong Chuang Zhang Zhanyu Ma Pattern Recognition and Intelligent Systems Lab. Beijing University of Posts and Telecommunications Beijing China PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education

Convolutional Neural Networks (CNNs) have been successfully applied in various image analysis tasks and gradually become one of the most powerful machine learning approaches. In order to improve the capability of the model generalization and performance in image classification, a new trend is to learn more discriminative features via CNNs. The main contribution of this paper is to increase the angles between the categories to extract discriminative features and enlarge the inter-class variance. To this end, we propose a loss function named focal inter-class angular loss (FICAL) which introduces the confusion rate-weighted cosine distance as the similarity measurement between categories. This measurement is dynamically evaluated during each iteration to adapt the model. Compared with other loss functions, experimental results demonstrate that the proposed FICAL achieved best performance among the referred loss functions on two image classification datasets.

关键词： Task analysis Training Feature extraction Computer vision Computational modeling Loss measurement Neural networks

来源：评论

学校读者我要写书评

暂无评论

Color image filtering based on impulsive noise detection

Color image filtering based on impulsive noise detection

引用

MIPPR 2007: Automatic Target recognition and Image Analysis;and Multispectral Image Acquisition

作者： Li, Dehua Jin, Lianghai Yao, Xingzhong Xiong, Caiquan State Key Lab. for Image Processing and Intelligent Control Institute for Pattern Recognition and Artificial Intelligence Huazhong University of Science and Technology Wuhan 430074 China Second Artillery Command College Wuhan 430012 China School of Computer Science Hubei University of Technology Wuhan 430068 China

ISBN: (纸本)9780819469502

This paper introduces a new class of switching vector median filter. The proposed algorithm first uses four directional masks to analyze the color difference between the central pixel and its neighborhood pixels in the RGB color space and classify each color pixel into noisy pixel or noise-free one, and then employs the standard vector median filtering operations in the detected noisy locations to restore the corrupted pixels and leave the noise-free ones unchanged. The simulation results show that the proposed method excellently suppresses impulsive noise as well as preserving the image details well, and significantly outperforms the existing vector filtering solutions in terms of both the objective measures and the perceptual visual quality.

关键词： Color image processing

来源：评论

学校读者我要写书评

暂无评论

DNN filter bank cepstral coefficients for spoofing detection

arXiv

引用

arXiv 2017年

作者： Yu, Hong Tan, Zheng-Hua Ma, Zhanyu Guo, Jun Pattern Recognition and Intelligent System Lab Beijing University of Posts and Telecommunications Beijing China Department of Electronic Systems Aalborg University Aalborg Denmark

With the development of speech synthesis techniques, automatic speaker verification systems face the serious challenge of spoofing attack. In order to improve the reliability of speaker verification systems, we develop a new filter bank based cepstral feature, deep neural network filter bank cepstral coefficients (DNN-FBCC), to distinguish between natural and spoofed speech. The deep neural network filter bank is automatically generated by training a filter bank neural network (FBNN) using natural and synthetic speech. By adding restrictions on the training rules, the learned weight matrix of FBNN is band-limited and sorted by frequency, similar to the normal filter bank. Unlike the manually designed filter bank, the learned filter bank has different filter shapes in different channels, which can capture the differences between natural and synthetic speech more effectively. The experimental results on the ASVspoof 2015 database show that the Gaussian mixture model maximum-likelihood (GMM-ML) classifier trained by the new feature performs better than the state-of-the-art linear frequency cepstral coefficients (LFCC) based classifier, especially on detecting unknown attacks. Copyright © 2017, The Authors. All rights reserved.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Directional Element HOG for Sketch recognition

Directional Element HOG for Sketch Recognition

引用

IEEE International Conference on Network Infrastructure and Digital Content (IC-NIDC)

作者： Yue Zhong Honggang Zhang Jun Guo Yi-Zhe Song Pattern Recognition and Intelligent System Lab Beijing University of Posts and Telecommunications Beijing China SketchX Lab Queen Mary University of London London United Kingdom

We propose a novel Directional Element Histogram of Oriented Gradient (DE-HOG) feature to human free-hand sketch recognition task that achieves superior performance to traditional HOG feature, originally designed for photographic objects. As a result of modeling the unique characteristics of free-hand sketch, i.e. consisting only a set of strokes omitting visual information such as color and brightness, being highly iconic and abstract. Specifically, we encode sketching strokes as a form of regularized directional vectors from the skeleton of a sketch, whilst still leveraging the HOG feature to meet the local deformation-invariant demands. Such a representation combines the best of two features by encoding necessary and discriminative stroke-level information, but can still robustly deal with various levels of sketching variations. Extensive experiments conducted on two large benchmark sketch recognition datasets demonstrate the performance of our proposed method.

关键词： Skeleton Feature extraction Task analysis Visualization Character recognition Histograms Image recognition

来源：评论

学校读者我要写书评

暂无评论

A Multi-homed VPN Architecture Based on Extended SOCKS+TLS Protocols

引用

Journal of Shanghai Jiaotong university(Science) 2005年第3期10卷 250-255页

作者：张宝良胡汉平陈翔 State Education Ministry Key Lab. of Image Information & Intelligent Control In st. of Pattern Recognition & Artificial Intelligence Huazhong Univ. of Science & Technology Wuhan 430074 China

A multi-homed VPN architecture based on extended SOCKSv5 and TLS was proposed. The architecture employs a dynamic connection mechanism for multiple proxies in the end system,i n which the security-demanded transmission connections can switch smoothly among the multiple proxies by maint aining a coherent connection *** mechanism is transparent to application programs and can support th e building of *** the cooperation of some other security components,the mechanism guarantees the reso urce availab.lity and reliability of the end system against some attacks to the specific ports or hosts.

关键词： multi-homed SOCKSv5 security connection context virtual private network (VPN)

来源：评论

学校读者我要写书评

暂无评论

Effect of multi-condition training and speech enhancement methods on spoofing detection

Effect of multi-condition training and speech enhancement me...

引用

International Workshop on Sensing, Processing and Learning for intelligent Machines (SPLINE)

作者： Hong Yu Achintya Sarkar Dennis Alexander Lehmann Thomsen Zheng-Hua Tan Zhanyu Ma Jun Guo Pattern Recognition and Intelligent System Lab. Beijing University of Posts and Telecommunications Dept. of Electronic Systems Aalborg University Aalborg Universitet Aalborg DK Beijing University of Posts and Telecommunications Beijing Beijing CN

ISBN: (纸本)9781467389181

Many researchers have demonstrated the good performance of spoofing detection systems under clean training and testing conditions. However, it is well known that the performance of speaker and speech recognition systems significantly degrades in noisy conditions. Therefore, it is of great interest to investigate the effect of noise on the performance of spoofing detection systems. In this paper, we investigate a multi-conditional training method where spoofing detection models are trained with a mix of clean and noisy data. In addition, we study the effect of different noise types as well as speech enhancement methods on a state-of-the-art spoofing detection system based on the dynamic linear frequency cepstral coefficients (LFCC) feature and a Gaussian mixture model maximum-likelihood (GMM-ML) classifier. In the experiment part we consider three additive noise types, Cantine, Babble and white Gaussian at different signal-to-noise ratios, and two mainstream speech enhancement methods, Wiener filtering and minimum mean-square error. The experimental results show that enhancement methods are not suitable for the spoofing detection task, as the spoofing detection accuracy will be reduced after speech enhancement. Multi-conditional training, however, shows potential at reducing error rates for spoofing detection.

关键词： Speech Speech enhancement Training Noise measurement Signal to noise ratio Databases Feature extraction

来源：评论

学校读者我要写书评

暂无评论

The comparison and analysis of tree data model and table-like data model based on NETCONF

The comparison and analysis of tree data model and table-lik...

引用

International Asia Conference on Informatics in Control, Automation and Robotics, CAR

作者： Yanxiang Gao Bin Zhang Guohui Li Yan Li Xuesong Gao Pattern Recognition and Intelligent System Lab Beijing University of Posts and Telecommunications Beijing China HuaWei Beijing Research Institute HuaWei Beijing Research Institute China

ISBN: (纸本)9781424451920

NETCONF is a new standard network configuration protocol. For further research we have developed an open experimental platform for NETCONF protocol, BUPT-NEP (BUPT NETCONF Experimental Platform). Moreover, based on the experimental platform some significant technologies, such as data filtering, data storage and so on, are considered. This paper gives a detailed introduction of the tree data model and the table-like data model, which both could be used to organize and store the network facilities' information, and highlights the comparison of them in two aspects of the space occupancy as well as the filtering performance. From the comparison results it comes to a conclusion that the table-like data model could keep more information; however, the tree data model is easy to understand and shows better filtering performance.

关键词： Data models Data analysis Protocols Information filtering Information filters Routing Robotics and automation intelligent robots Informatics Automatic control

来源：评论

学校读者我要写书评

暂无评论

Mobile big data analysis with machine learning

arXiv

引用

arXiv 2018年

作者： Xie, Jiyang Song, Zeyu Li, Yupeng Ma, Zhanyu Pattern Recognition and Intelligent Systems Lab. Beijing University of Posts and Telecommunications Beijing China State Key Lab. of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

This paper investigates to identify the requirement and the development of machine learning-based mobile big data analysis through discussing the insights of challenges in the mobile big data (MBD). Furthermore, it reviews the state-of-the-art applications of data analysis in the area of MBD. Firstly, we introduce the development of MBD. Secondly, the frequently adopted methods of data analysis are reviewed. Three typical applications of MBD analysis, namely wireless channel modeling, human online and offline behavior analysis, and speech recognition in the internet of vehicles, are introduced respectively. Finally, we summarize the main challenges and future development directions of mobile big data analysis. Copyright © 2018, The Authors. All rights reserved.

关键词： Big data

来源：评论

学校读者我要写书评

暂无评论

Cross-layer navigation convolutional neural network for fine-grained visual classification

arXiv

引用

arXiv 2021年

作者： Guo, Chenyu Xie, Jiyang Liang, Kongming Sun, Xian Ma, Zhanyu Pattern Recognition and Intelligent System Lab Beijing University of Posts and Telecommunications Beijing China Aerospace Information Research Institute Chinese Academy of Sciences Beijing China

Fine-grained visual classification (FGVC) aims to classify sub-classes of objects in the same super-class (e.g., species of birds, models of cars). For the FGVC tasks, the essential solution is to find discriminative subtle information of the target from local regions. Traditional FGVC models preferred to use the refined features, i.e., high-level semantic information for recognition and rarely use low-level information. However, it turns out that low-level information which contains rich detail information also has effect on improving performance. Therefore, in this paper, we propose cross-layer navigation convolutional neural network for feature fusion. First, the feature maps extracted by the backbone network are fed into a convolutional long short-term memory model sequentially from high-level to low-level to perform feature aggregation. Then, attention mechanisms are used after feature fusion to extract spatial and channel information while linking the high-level semantic information and the low-level texture features, which can better locate the discriminative regions for the FGVC. In the experiments, three commonly used FGVC datasets, including CUB-200-2011, Stanford-Cars, and FGVC-Aircraft datasets, are used for evaluation and we demonstrate the superiority of the proposed method by comparing it with other referred FGVC methods to show that this method achieves superior *** Codes 14J60 (Primary) 14F05, 14J26 (Secondary) Copyright © 2021, The Authors. All rights reserved.

关键词： Convolution

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：