Mining Relations Among Cross-Frame Affinities for Video Semantic Segmentation.
Guolei SunYun LiuHao TangAjad ChhatkuliLe ZhangLuc Van GoolPublished in: ECCV (34) (2022)
Keyphrases
- semantic segmentation
- video frames
- key frames
- conditional random fields
- superpixels
- label transfer
- scene classification
- weakly supervised
- video data
- video sequences
- object categories
- pascal voc
- object detection
- object class
- multiscale
- pairwise
- object classes
- object recognition
- long range
- text mining
- saliency map
- denoising
- spatial information
- text classification