SlowFast Rolling-Unrolling LSTMs for Action Anticipation in Egocentric Videos.
Nada OsmanGuglielmo CamporesePasquale CosciaLamberto BallanPublished in: ICCVW (2021)
Keyphrases
- human actions
- activity recognition
- action classification
- action recognition
- recognizing actions
- recognizing human actions
- video sequences
- video summarization
- video frames
- human activities
- user generated
- video data
- action detection
- video surveillance
- recognition of human actions
- event recognition
- view invariant
- space time interest points
- video database
- video content
- multimedia
- action selection
- moving camera
- motion features
- action sequences
- key frames
- video streams
- spatio temporal
- moving objects
- neural network
- real time