4 ms·A large-scale dataset for recognizing and understanding action in videos1 points by interconnector 8y ago