An automatic face indexing framework for actor-based video services in an IPTV environment

Automatic face indexing is a promising technique for personalized IPTV applications. This paper proposes a novel framework that makes use of an STB's Internet connection in order to realize face indexing for a personal media library. In particular, given a drama or movie title, actor names are first retrieved from a drama or movie information web service. The obtained actor names are subsequently used to acquire training and gallery face images from the Internet. More than 20,000 video frames collected from five drama video sequences were used to evaluate our face indexing framework. Experimental results show that the obtained face indexing accuracy is promising for realizing actor-based video services in an IPTV environment.

[1]  Yong Man Ro,et al.  Color Face Recognition for Degraded Face Images , 2009, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[2]  Yong Man Ro,et al.  Face annotation for personal photos using context-assisted face recognition , 2008, MIR '08.

[3]  Yong Man Ro,et al.  Meaningful scene filtering for TV terminals , 2006, 2006 Digest of Technical Papers International Conference on Consumer Electronics.

[4]  Tieniu Tan,et al.  Online Appearance Model Learning for Video-Based Face Recognition , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[5]  Juan F. Sevillano,et al.  An approach to simplify the design of IFFT/FFT cores for OFDM systems , 2006, IEEE Transactions on Consumer Electronics.