SASFormer: Transformers for Sparsely Annotated Semantic Segmentation.
Hui SuYue YeWei HuaLechao ChengMingli SongPublished in: CoRR (2022)
Keyphrases
- semantic segmentation
- street scenes
- image set
- label transfer
- conditional random fields
- superpixels
- weakly supervised
- scene classification
- object categories
- object classes
- manually annotated
- object class
- image understanding
- pascal voc
- long range
- image processing
- named entities
- sparse representation
- segmentation algorithm
- face images
- information extraction
- pairwise