Objects2action: Classifying and Localizing Actions without Any Video Example.

Mihir Jain, Jan C. van Gemert, Thomas Mensink, Cees G. M. Snoek

Browse the full ICCV paper archive.