【文章內(nèi)容簡(jiǎn)介】 頻看成一個(gè)整體來(lái)克服分辨率低的問(wèn)題[35]. 對(duì)于光照或者姿態(tài)的單獨(dú)變化可以通過(guò)矩陣、概率或者流形的方式部分解決[9, 18], 但是需要不同條件下的大量的訓(xùn)練樣本. 對(duì)于遮擋問(wèn)題可以采用魯棒統(tǒng)計(jì)學(xué)[11]或者對(duì)臉部的分塊處理[54]來(lái)解決.隨著研究的深入, 基于視頻的人臉識(shí)別需要進(jìn)一步研究的工作包括:(1) 人臉特征的準(zhǔn)確定位本文假設(shè)已經(jīng)得到了圖像或者視頻中人臉的位置, 并且人臉的特征已經(jīng)準(zhǔn)確定位. 但是在實(shí)際應(yīng)用中, 人臉視頻的分辨率過(guò)低常會(huì)使得人臉的檢測(cè)和準(zhǔn)確的特征定位存在一定的困難. 人臉的誤配準(zhǔn)也會(huì)嚴(yán)重影響人臉識(shí)別的結(jié)果. 作為人臉識(shí)別的基礎(chǔ), 準(zhǔn)確和快速的人臉檢測(cè)和特征定位方法是必不可少的.(2) 人臉的超分辨率重建和模糊復(fù)原視頻序列中的人臉由于采集條件和運(yùn)動(dòng)的影響, 人臉圖像分辨率低且人臉模糊. 研究人臉圖像超分辨率技術(shù)[55]和圖像復(fù)原技術(shù)[56]以得到清晰的人臉圖像也是未來(lái)需要重點(diǎn)解決的問(wèn)題.(3) 人臉的3D建?，F(xiàn)階段基于二維的人臉識(shí)別方法可以在一定程度上解決姿態(tài)或光照的變化問(wèn)題. 但是人臉是一個(gè)三維的物體, 利用人臉的三維的信息是解決姿態(tài), 光照變化問(wèn)題的最本質(zhì)方法. 現(xiàn)階段利用視頻數(shù)據(jù)生成3D模型的計(jì)算復(fù)雜度很大[42, 5759], 無(wú)法達(dá)到使用要求. 更好降低三維人臉建模的復(fù)雜度和提高建模的精度是未來(lái)發(fā)展的一個(gè)重要方向.(4) 視頻人臉數(shù)據(jù)庫(kù)和測(cè)試方法的標(biāo)準(zhǔn)化與基于靜止圖像的人臉識(shí)別相比, 基于視頻的人臉識(shí)別的最大問(wèn)題是還沒(méi)有一個(gè)包含各種條件變化的、統(tǒng)一的、大規(guī)模的視頻人臉數(shù)據(jù)庫(kù)和測(cè)試標(biāo)準(zhǔn). 許多文章采用的視頻人臉數(shù)據(jù)庫(kù)和測(cè)試方法都不盡相同, 無(wú)法進(jìn)行算法之間的比較. 建立一個(gè)公共的、大規(guī)模的視頻人臉數(shù)據(jù)庫(kù)和標(biāo)準(zhǔn)的測(cè)試方法是該領(lǐng)域的一個(gè)首要任務(wù).(5) 多模生物特征認(rèn)證現(xiàn)階段基于視頻的人臉識(shí)別算法主要是基于室內(nèi)的環(huán)境條件. 室外條件下的人臉圖像光照、姿態(tài)等的劇烈變化使人臉識(shí)別仍然面臨著許多困難, 融合多種生物特征提高識(shí)別的性能也將是未來(lái)研究的一個(gè)重點(diǎn)[6062].參考文獻(xiàn)[1] Chellappa R, Wilson C, Sirohey S. Human and machine recognition of faces: A survey. Proceedings of the IEEE, 1995, 83(5): 705740[2] Zhao W, Chellappa R, Rosenfeld A, Phillips P J. Face recognition: A literature survey. ACM Computation Survey, 2003, 35(4): 399458[3] Li S Z, Jain A K. Handbook of Face Recognition, Springer, New York, 2005[4] Zhou S, Chellappa R. Beyond a single still image: face recognition from multiple still images and videos. Face Processing: Advanced Modeling and Methods, Academic Press, New York, 2005[5] Shakhnarovich G., Fisher J W, Darrell T. Face recognition from longterm observations//Proceedings of the European Conference on Computer Vision. Bari, 2002: 851868[6] Liu X M, Chen T, Thornton S M. Eigenspace updating for nonstationary process and its application to face recognition. Pattern Recognition, 2003, 36(9): 19451959[7] Liu X M, Chen T. Videobased face recognition using adaptive hidden markov models//Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition. Madison, 2003: 340345[8] Lee K C, Ho J, Yang M H, Kriegman D. Videobased face recognition using probabilistic appearance manifolds//Proceedings of the International IEEE Conference on Computer Vision and Pattern Recognition. Madison, 2003: 313320[9] Lee K C, Ho J, Yang M H, Kriegman D. Visual tracking and recognition using probabilistic appearance manifolds. Computer Vision and Image Understanding, 2005, 99(3): 303331[10] Zhou S, Krueger V, Chellappa R. Probabilistic recognition of human faces from video. Computer Vision and Image Understanding, 2003, 91(1): 214245[11] Zhou S, Chellappa R, Moghaddam B. Visual tracking and recognition using appearanceadaptive models in particle filters. IEEE Transactions on Image Processing, 2004, 13(11): 14341456[12] Aggarwal G., Chowdhury A K R, Chellappa R. A system identification approach for videobased face recognition//Proceedings of the IEEE International Conference on Pattern Recognition. Cambridge, 2004: 2326[13] Arandjelovi? O, Cipolla R. Face recognition from face motion manifolds using robust kernel resistoraverage distance//Proceedings of the IEEE Conference on Compute Vision and Patter Recognition workshop. Washington , 2004: 8893[14] Arandjelovi? O, Shakhnarovich G, Fisher G, Cipolla R, Darrell T. Face recognition with image sets using manifold density divergence//Proceeding of the IEEE Conference on Computer Vision and Pattern Recognition. San Diego, 2005: 581588[15] Arandjelovi? O, Cipolla R. A posewise linear illumination manifold model for face recognition using video. Computer Vision and Image Understanding, 2009, 113(1): 113125[16] Yamaguchi O, Fukui K, Maeda K. Face recognition using temporal image sequence//Proceedings of the IEEE International Conference on Automatic Face and Gesture Recognition. Nara, 1998: 318323[17] Fukui K, Yamaguchi O. Face recognition using multiviewpoint patterns for robot vision. International Symposium of Robotics Research. 2003: 192–201.[18] Nishiyama M, Yamaguchi O, Fukui K. Face Recognition with the multiple constrained mutual subspace method//Proceeding of the 5th International Conference on Audio and VideoBased Biometric Person Authentication. New York, 2005: 7180[19] Li J W, Wang Y H, Tan T N. Videobased face recognition using a metric of average Euclidean distance// Proceedings of the 5th Chinese Conference on Biometric Recognition. Guangzhou, 2004: 224232[20] Li J W, Wang Y H, Tan TN. Videobased face recognition using earth mover’s distance//Proceedings of the International Conference on Audio and Videobased person authentication. New York, 2005: 229239[21] Fan W, Wang Y H, Tan T N. Videobased face recognition using Bayesian inference model//Proceedings of the International Conference on Audio and Videobased person authentication. New York, 2005: 122130[22] Yan Y, Zhang Y J. Stateoftheart on videobased face recognition. Encyclopedia of Artificial Intelligence, 2008, 14551461[23] Jia H X, Zhang Y J. Human detection in static images//Verma B, Blumenstein M. Pattern Recognition Technologies and Applications: Recent Advances. 2008, 227243[24] Liu X M, Zhang Y J, Tan H C. A new hausdorff distance based approach for face localization. Sciencepaper Online,