Learning Self-Similarity in Space and Time as Generalized Motion for Video Action Recognition.

Heeseung Kwon, Manjin Kim, Suha Kwak, Minsu Cho

Browse the full ICCV paper archive.