Weakly-supervised Disentanglement Network for Video Fingerspelling Detection.
Ziqi JiangShengyu ZhangSiyuan YaoWenqiao ZhangSihan ZhangJuncheng LiZhou ZhaoFei WuPublished in: ACM Multimedia (2022)
Keyphrases
- weakly supervised
- object detectors
- weakly labeled
- object class
- relation extraction
- object detection
- topic models
- video data
- superpixels
- semi supervised
- named entities
- partial occlusion
- multiscale
- information retrieval
- domain specific
- natural images
- wordnet
- machine learning
- video frames
- natural language
- image processing
- learning algorithm