Three-stream spatio-temporal attention network for first-person action and interaction recognition
Keyword(s):
Keyword(s):
Keyword(s):
Keyword(s):