A large-scale dataset for recognizing and understanding action in videos

1 points| interconnector | 8 years ago |moments.csail.mit.edu