检索结果-南通市图书馆

Chinese Word Recognition and Understanding with Information Feedback

在线全文

学校读者我要写书评

暂无评论

Chinese Word Recognition and Understanding with Information ...

1996 3rd International Conference on Signal Processing(ICSP’96)

作者： Fu Qiuliang yuan baozong Lin Biqin Institute of Information Science Northern Jiaotong University

ISBN: (纸本)0780329120

＜正＞It is the one of tasks of Chinese speech understanding system that translated Chinese syllable that is produced by Chinese speech recognition system into a correct Chinese word,and Chinese speech understanding system will implement the task of transform from speech to text altogether with Chinese speech recognition *** paper make a system on Chinese word recognition and understanding on a new approach with information feedback, different from traditional scheme of speech recognition and understanding,they are both linguistic method and statistical ***,this paper gives a conclusion on a experimental system and summary about the Chinese speech recognition and understanding approach with information feedback.

关键词： speech synthesis speech recognition feedback natural languages statistical analysis linguistics Chinese word recognition Chinese word understanding information feedback Chinese syllable Chinese speech recognition system statistical methods linguistic methods experimental system text to speech synthesis

Markerless human motion capture by Markov random field and dynamic graph cuts with color constraints

在线全文

cnki会议

学校读者我要写书评

暂无评论

Science in China(Series F) 2009年第2期52卷 252-259页

作者： LI Jia WAN ChengKai ZHANG DianYong MIAO ZhenJiang yuan baozong Institute of Information Science Beijing Jiaotong University Beijing 100044 China

Currently, many vision-based motion capture systems require passive markers attached to key loca- tions on the human body. However, such systems are intrusive with limited application. The algorithm that we use for human motion capture in this paper is based on Markov random field （MRF） and dynamic graph cuts. It takes full account of the impact of 3D reconstruction error and integrates human motion capture and 3D reconstruction into MRF-MAP framework. For more accurate and robust performance, we extend our algorithm by incorporating color constraints into the pose estimation process. The advantages of incorporating color constraints are demonstrated by experimental results on several video sequences.

关键词： human motion capture Markov random field dynamic graph-cut color constrains

FACE RECOGNITION USING TWO DIMENSIONAL LAPLACIAN EIGENMAP

在线全文

学校读者我要写书评

暂无评论

Journal of Electronics(China) 2008年第5期25卷 616-621页

作者： Chen Jiangfeng yuan baozong Pei Bingnan Institute of Information Science Bcijing Jiaotong University Beijing 100044 China School of Information Engineering Dalian University Dalian 116522 China

Recently, some research efforts have shown that face images possibly reside on a nonlinear sub-manifold. Though Laplacianfaees method considered the manifold structures of the face images, it has limits to solve face recognition problem. This paper proposes a new feature extraction method, Two Dimensional Laplacian EigenMap （2DLEM）, which especially considers the manifold structures of the face images, and extracts the proper features from face image matrix directly by using a linear transformation. As opposed to Laplacianfaces, 2DLEM extracts features directly from 2D images without a vectorization preprocessing. To test 2DLEM and evaluate its performance, a series of ex- periments are performed on the ORL database and the Yale database. Moreover, several experiments are performed to compare the performance of three 2D methods. The experiments show that 2DLEM achieves the best performance.

关键词： (2DLEM) Face recognition Manifold Laplacianfaces Two Dimensional Laplacian EigenMap

ICSP’93 Technical Program Overview

在线全文

学校读者我要写书评

暂无评论

ICSP’93 Technical Program Overview

International Conference on Signal Processing(ICSP’93)

作者： yuan baozong Technical Program Chairman

＜正＞It is a pleasure for me to welcome you all to the historic city Beijing to attend a modern technical conference,the Second International Conference on Signal Processing ICSP’*** the three years,since the first International conference ICSP was held in 1990,there has been a substantial progress in theory of digital signal processing and its *** Processing has become one of the basic technologies in developing modern information industries and social *** is the reason why the ICSP got so overwhelming response for their call for papers from countries around the *** year we received 458

关键词：

Approach to Human-machine Natural Interactive System

在线全文

cnki会议

学校读者我要写书评

暂无评论

Approach to Human-machine Natural Interactive System

1998 Fourth International Conference Signal Processing

作者： Chen Fang yuan baozong Institute of Information Science Northern Jiaotong University

＜正＞The paper introduced a human-machine natural interactive system which based on Chinese understanding and Chinese *** system can accept the question in the form of natural language and generate the description of... 详细信息

关键词： Chinese understanding Chinese generation knowledge inquiring reasoning learning

A NEW EDGE DETECTOR WITH THINNING AND NOISE RESISTING ABILITIES

在线全文

cnki会议

学校读者我要写书评

暂无评论

Journal of Electronics(China) 1989年第4期6卷 314-319页

作者：陈根明袁保宗 Northern Jiaotong University Beijing

This paper presents a new edge detector using 5 ~ 5 mask, which can reduce noise efficientlybut not increase the width of the detected edge which always happens in the case of using the 5×5 windowedge detector. Besides, in order to handle the problem that the contrast decreases in the dark regioncaused by underexposure, this detector uses a self-adjusting threshold, so that it can detect the edge in re-gions of different grey background correctly.

关键词： Digital image processing Edge detection

A LINEAR ALGORITHM FOR 3-D MOTION ESTIMATION BASED ON THE ORIENTATION INVARIANCE OF ROTATIONAL AXIS

在线全文

学校读者我要写书评

暂无评论

A LINEAR ALGORITHM FOR 3-D MOTION ESTIMATION BASED ON THE OR...

International Conference on Signal Processing(ICSP’90)

作者： Fan Hong yuan baozong Institute of Information Science Northern Jiaotong University

＜正＞This paper proposes a linear algorithm for determining 3-D motion parameters of a rigid ***,we solve for an intermediate parameter matrix from more than 8 image point correspondences of two views,and then compute the translation components up to a scale factor from a set of linear *** on the orientation invariance of the rotational axis, the axis orientation and rotational angle can be solved by a set of linear *** the physical restriction,we show a criterion to determine the unique solution in one computation.

关键词：

A D-S Based Multi-Channel Information Fusion Method Using Classifier’s Uncertainty Measurement

在线全文

cnki会议

学校读者我要写书评

暂无评论

A D-S Based Multi-Channel Information Fusion Method Using Cl...

2000年第16届世界计算机大会

作者： Liu Rujie yuan baozong The Institute of Information Science Northern JiaoTong Univ.

＜正＞Classifiers of different types complement one another in classification performance,and there exists uncertainty in the outputs of classifiers,which reflects the reliability of the classification results. However,this uncertainty is not considered in traditional fusion *** this paper,a geometrically inspired measurement of uncertainty is proposed which is integrated when combining multiple classifiers using D-S evidence theory. Numerical analysis results show that this method is effective and reasonable.

关键词： D-S Method Information Fusion

MODELING OF 3D CURVED OBJECTS FROM MONOCULAR IMAGES

在线全文

cnki会议

学校读者我要写书评

暂无评论

MODELING OF 3D CURVED OBJECTS FROM MONOCULAR IMAGES

International Conference on Signal Processing(ICSP’93)

作者： Zhang Chao yuan baozong Institute of Information Science Northern Jiaotong University

＜正＞This paper presents a method for the generation of 3D curved models from monocular *** models,represented by deformed superquadrics, are generated from contours on orthographic planes of intensity images taken from known positions. The parameters of the models are determined fully automatically by an iterative *** shape adaption for each image includes a global fitting where the silhouttes of the object detected by an edge detector are fitted to the parameteric model,a local fitting where some deformation parameters are adjusted to the detected *** these parameters,the geometric shape descriptions of 3D objects are obtained.

关键词：