Abstract
A method for modelling and recognising facial identity is described within the context of an integrated system for face recognition in dynamic scenes. Recognition is based sequences rather than isolated images. Mixture models provide estimates of class-conditional probabilities and these are used to accumulate recognition confidence over time. Results are presented using data from the integrated system.
Cite
CITATION STYLE
McKenna, S. J., & Gong, S. (1997). Face recognition from sequences using models of identity. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1351, pp. 507–514). Springer Verlag. https://doi.org/10.1007/3-540-63930-6_161
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
