A proposal for high compression of faces in video sequences using adaptive eigenspaces

The paper presents a proposal for a novel video coding scheme intended to encode human faces in video sequences at very high compression using a recognition and reconstruction approach. The scheme is based on the well-known eigenspace concepts used in face recognition systems, which have been modified to cope with the video compression application. An adaptive mechanism is presented to update the eigenspace which improves the overall approach. The presented first results look promising.

[1]  Edward J. Delp,et al.  New trends in image and video compression , 2000, 2000 10th European Signal Processing Conference.

[2]  Alex Pentland,et al.  Probabilistic Visual Learning for Object Representation , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  Iso/iec 14496-2 Information Technology — Coding of Audio-visual Objects — Part 2: Visual , 2022 .

[4]  Peter Eisert,et al.  Analyzing Facial Expressions for Virtual Conferencing , 1998, IEEE Computer Graphics and Applications.

[5]  Fernando Pereira Visual data representation: Recent achievements and future developments , 2000, 2000 10th European Signal Processing Conference.

[6]  H. Harashima,et al.  Model-Based Analysis Synthesis Coding of Videotelephone Images--Conception and Basic Study of Intelligent Image Coding-- , 1989 .

[7]  Verónica Vilaplana,et al.  Human face segmentation and tracking using connected operators and partition projection , 1999, Proceedings 1999 International Conference on Image Processing (Cat. 99CH36348).

[8]  Luis Torres,et al.  Automatic face recognition for video indexing applications , 2002, Pattern Recognit..

[9]  James L. Crowley,et al.  Face-Tracking and Coding for Video Compression , 1999, ICVS.

[10]  Faouzi Kossentini,et al.  H.263+: video coding at low bit rates , 1998, IEEE Trans. Circuits Syst. Video Technol..