Weakly-Supervised Video Object Grounding by Exploring Spatio-Temporal Contexts.
Xun YangXueliang LiuMeng JianXinjian GaoMeng WangPublished in: ACM Multimedia (2020)
Keyphrases
- weakly supervised
- video objects
- spatio temporal
- object class
- superpixels
- video sequences
- motion trajectories
- image sequences
- topic models
- relation extraction
- object segmentation
- object tracking
- semi supervised
- named entities
- object recognition
- moving objects
- video frames
- object detection
- human motion
- video content
- key frames
- co occurrence
- high resolution
- viewpoint
- appearance model
- bag of words
- machine learning