一种融合流形学习的视频人脸性别识别改进算法

如何有效利用视频中人脸之间的时空连续性信息来克服人脸分辨率低、图像尺度变化大和姿态、光照变化以及遮挡等问题是视频人脸识别的关键所在.提出了一种基于流形学习的视频人脸性别识别算法.该算法不仅可以通过聚类融合学习来挖掘视频内在的连续性信息,同时能发现人脸数据中内在非线性结构信息而获得低维本质的流形结构.在UCSD/Honda和自采集数据库上与静态的算法比较结果表明,所提算法能够获得更好的识别率.

An improved manifold-based face gender recognition algorithm for video

How to fully utilize both spatial and temporal information in video to overcome the difficulties existing in the video-based 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,has become the research focus.In this paper,a novel manifold-based face gender recognition algorithm for video(VG-LPP) using clustering is proposed,which can discover more special semantic information hidden in video face sequence,simultaneously well utilize the intrinsic nonlinear structure information to extract discriminative manifold features.Comparison of VG-LPP with other algorithms on UCSD/Honda and the author′s own video databases shows that the proposed approach can perform better for video-based face gender recognition.