Watching Unlabeled Video Helps Learn New Human Actions from Very Few Labeled Snapshots.

Chao-Yeh Chen, Kristen Grauman

Browse the full CVPR paper archive.