Audio-visual Person Veriication Audio-visual Person Veriication

In this paper we investigate bene ts of classi er combination fusion for a multimodal system for personal identity veri cation The system uses frontal face images and speech We show that a sophisticated fusion strategy enables the system to outperform its facial and vocal modules when taken seperately We show that both trained linear weighted schemes and fusion by Support Vector Machine classi er leads to a signi cant reduction of total error rates The complete system is tested on data from a publicly available audio visual database XM VTS subjects according to a published protocol