Weakly-Supervised Video Moment Retrieval via Semantic Completion Network.
Zhijie LinZhou ZhaoZhu ZhangQi WangHuasheng LiuPublished in: CoRR (2019)
Keyphrases
- weakly supervised
- semantic attributes
- weakly labeled
- relation extraction
- topic models
- video data
- video sequences
- natural language
- information retrieval
- superpixels
- image database
- semantic information
- semantic similarity
- markov random field
- named entities
- image content
- video frames
- semantic features
- semi supervised
- viewpoint