Small Objects Matters in Weakly-supervised Semantic Segmentation.
Cheolhyun MunSanghuk LeeYoungjung UhJunsuk ChoeHyeran ByunPublished in: CoRR (2023)
Keyphrases
- weakly supervised
- semantic segmentation
- superpixels
- object classes
- object class
- object segmentation
- bounding box
- object detectors
- relation extraction
- multiple objects
- object categories
- topic models
- conditional random fields
- multiscale
- semi supervised
- d objects
- named entities
- small number
- object recognition
- object model
- computer vision
- pascal voc
- target object
- image features
- information extraction
- bayesian networks
- scene classification
- viewpoint
- object detection
- partial occlusion
- image regions