CDC: Convolutional-De-Convolutional Networks for Precise Temporal Action Localization in Untrimmed Videos.
Zheng ShouJonathan ChanAlireza ZareianKazuyuki MiyazawaShih-Fu ChangPublished in: CVPR (2017)
Keyphrases
- deep learning
- human actions
- network architecture
- action recognition
- video sequences
- motion history images
- temporal information
- np complete
- recognizing human actions
- social networks
- deep belief networks
- action classification
- temporal relationships
- spatial and temporal
- sparse coding
- temporal data
- spatio temporal
- video frames
- video surveillance
- temporal segmentation
- temporal order
- community structure
- atomic actions
- human activities
- complex networks