Video to the Rescue

Chetty, G. and Wagner, M.

Automatic person identity verification based on biometrics is a challenging problem, and has received much attention during recent years due to its many applications in on-line transaction processing, law enforcement, and security applications. However, most identity verification systems are primarily based on voice biometrics, and hence are more vulnerable to acoustic noise and channel distortions, in addition to train/test mismatch conditions. In this paper, we show how we can use video information to improve the performance of identity verification systems. The approach based on multimodal fusion of voice and face information from speaking face video allows robust identity verification performance. In addition, depending on the type of features and fusion technique used, it is also possible to perform liveness checks, allowing the system to detect fraudulent attacks on the system.

Cite as: Chetty, G. and Wagner, M. (2006). Video to the Rescue. In Proc. HCSNet Workshop on the Use of Vision in Human-Computer Interaction, (VisHCI 2006), Canberra, Australia. CRPIT, 56. Goecke, R., Robles-Kelly, A. and Caelli, T., Eds. ACS. 107-108.

(from crpit.com) (local if available)