SLRNet: Semi-Supervised Semantic Segmentation Via Label Reuse for Human Decomposition Images.
Sara MousaviZhenning YangKelley CrossDawnie W. SteadmanAudris MockusPublished in: CoRR (2022)
Keyphrases
- semantic segmentation
- street scenes
- semi supervised
- label transfer
- ground truth
- input image
- image database
- image understanding
- three dimensional
- image features
- object recognition
- superpixels
- weakly supervised
- image classification
- semi supervised learning
- similarity measure
- pixel level
- image set
- image regions
- segmentation algorithm
- image annotation
- supervised learning
- spatial information
- partial occlusion
- multiple images
- bounding box
- unlabeled data
- labeled data
- natural images
- multi view
- object detection