Rethinking Segmentation Guidance for Weakly Supervised Object Detection.
Ke YangPeng ZhangPeng QiaoZhiyuan WangHuadong DaiTianlong ShenDongsheng LiYong DouPublished in: CVPR Workshops (2020)
Keyphrases
- weakly supervised
- object detection
- object class
- semantic segmentation
- object segmentation
- weakly supervised learning
- superpixels
- object detectors
- fully supervised
- segmentation algorithm
- object categories
- segmentation method
- image segmentation
- multiscale
- topic models
- object recognition
- medical images
- bounding box
- shape prior
- graph cuts
- computer vision
- object classes
- multi class
- viewpoint
- relation extraction
- named entities
- level set
- multiple objects
- image processing
- probabilistic model
- image regions
- automatic extraction
- video sequences
- information retrieval
- energy function