【正文】 CMURITR0118, Robotics Institute, Carnegie Mellon University, 2001[53] Teferi D, Bigun J. Damascening video databases for evaluation of face tracking and recognition The DXM2VTS database. Pattern Recognition Letters, 2007, 28(15): 21432156[54] Zhang Y, Martinez A M. A weighted probabilistic approach to face recognition from multiple images and video sequences. Image and Vision Computing, 2006, 24(6): 626638[55] AlAzzeh M, Eleyan A, Demirel H. PCAbased face recognition from video using superresolution. Proceedings of 23rd International Symposium on Computer and Information Sciences, Istanbul, 2008, 14[56] Goksel D. Exploiting spacetime statistics of videos for face hallucination. . Thesis, Carnegie Mellon University, 2007.[57] Chowdhury A, Chellappa R. Face reconstruction from monocular video using uncertainty analysis and a generic model. Computer Vision and Image Understanding, 2003, 91(1): 188213[58] Choudhury A, Clarkson B, Jebara T, Penland A. Multimodal person recognition using unconstrained audio and video//Proceedings of the conference on Audio and Videobased Biometric Person Authentication, Washington , 1999: 176180[59] Zhang Z Y, Liu Z C, Adler D, Cohen M F, Hanson E, Shan Y. Robust and rapid generation of animated faces from video images: a modelbased modeling approach. International Journal of Computer Vision, 2004, (58)2: 93119[60] Zhou X, Bhanu B. Integrating face and gait for human recognition at a distance in video. IEEE Transactions on Systems, Man and Cybernetics, Part B, 2007, 37(5): 11191137[61] Jing X Y, Yao Y F, Zhang D, Yang J Y, Li M. Face and palmprint pixel level fusion and kernel DCVRBF classifier for small sample biometric recognition. Pattern Recognition, 2007, 40(11): 32093324.[62] Yan Y, Zhang Y J. Multimodal biometrics fusion using correlation filter bank//Proceedings of 19th IAPR International Conference on Pattern Recognition, Tampa, 2008, (14)StateoftheArt on VideoBased Face RecognitionYAN Yan1), 2) ZHANG YuJin1), 2)1) (Tsinghua National Laboratory for Information Science and Technology, Tsinghua University, Beijing 100084)2) (Department of Electronic Engineering, Tsinghua University, Beijing 100084)Abstract Recently, videobased face recognition has bee one of the hottest topics in the domain of face recognition. How to fully utilize both spatial and temporal information in video to overe the difficulties existing in the videobased face recognition, such as low resolution of face images in video, large variations of face scale, radical changes of illumination and pose as well as occasionally occlusion of different parts of faces, is the focus. The paper reviews most existing typical methods for videobased face recognition (especially for the last 5 years) and analyses their respective pros and cons. Two monly used video face databases and some experimental results are given. The prospects for future development and suggestions for further research works are put forward in the end.Keywords Pattern recognition, Face recognition, Videobased face recognition, Progress, SurveyBackgroundThis work is supported by the National Natural Science Foundation of China under Grant No. 60872084 and the Specialized Research Fund for the Doctoral Program of Higher Education under Grant No. 20060003102.Traditional still imagebased face recognition has achieved great success in constrained environments. However, once the conditions (including illumination, pose, expression, age, etc.) change too much, the performance declines dramatically. The recent FRVT2002 shows that the recognition performance of face images captured in an outdoor environment and different days is still not satisfying. Current still imagebased face recognition algorithms are even far away from the capability of human perception system. On the other hand, psychology and physiology studies have shown that motion can help people for better face recognition. During the past several years, many research efforts have been concentrated on videobased face recognition. Compared with still imagebased face recognition, true videobased face recognition algorithms that use both spatial and temporal information started only a few years ago. No prehensive survey in this field has been made, and a lot of issues in videobased face recognition still have not been addressed well. So the content of this paper gives an overview of the most existing methods in the field of videobased face recognition. A suitable classification for different methods has been made, the respective pros and cons of typical techniques in each method group are analysed. The important issues which need to be solved, the prospects for future development and some suggestions for further research works are put forward to meet the goal of this paper. 作者簡(jiǎn)歷（中文見第1頁題注）Yan Y, born in 1984, , main research area is pattern recognition. Zhang YJ, born in 1954, Professor, main research area is image engineering (image processing, image analysis, image understanding and technique application), ~zhangyujin/第一作者照片聯(lián)系人電話和Email聯(lián)系人：章毓晉；電話：62781430；Email: zhangyj@12