A combination of generative and discriminative models for fast unsupervised activity recognition from traffic scene videos.

Mahesh Venkata Krishna, Joachim Denzler

Browse the full WACV paper archive.