3D Self-Attention for Unsupervised Video Quantization.
Jingkuan SongRuimin LangXiaosu ZhuXing XuLianli GaoHeng Tao ShenPublished in: SIGIR (2020)
Keyphrases
- video sequences
- video data
- real time
- video content
- video frames
- video streams
- unsupervised manner
- multimedia
- unsupervised learning
- video processing
- semi supervised
- real time video
- space time
- spatial and temporal
- video segmentation
- online video
- visual saliency
- machine learning
- quantization error
- video images
- video database
- data driven
- saliency map
- video clips
- video retrieval
- visual attention
- human activities
- spatio temporal