Maximizing Spatio-Temporal Entropy of Deep 3D CNNs for Efficient Video Recognition.
Junyan WangZhenhong SunYichen QianDong GongXiuyu SunMing LinMaurice PagnuccoYang SongPublished in: CoRR (2023)
Keyphrases
- spatio temporal
- spatial and temporal
- spatial temporal
- space time
- human actions
- video data
- multimedia
- video streams
- action recognition
- mutual information
- feature extraction
- pattern recognition
- recognition rate
- object recognition
- human activities
- video content
- real time
- video representation
- spatio temporally
- information theoretic
- video sequences
- recognition accuracy
- video surveillance
- information theory
- automatic recognition
- recognition process
- temporal domain