Few-shot semantic segmentation via mask aggregation.
Wei AoShunyi ZhengYan MengPublished in: CoRR (2022)
Keyphrases
- semantic segmentation
- street scenes
- conditional random fields
- superpixels
- label transfer
- scene classification
- weakly supervised
- pascal voc
- bounding box
- object classes
- object detection
- object categories
- video data
- video sequences
- object class
- key frames
- segmentation algorithm
- visual features
- natural images
- multiscale
- image sequences
- natural scenes
- image understanding
- image segmentation
- n gram
- image representation
- training data