Weakly-Supervised Video Object Grounding from Text by Loss Weighting and Object Interaction.
Luowei ZhouNathan LouisJason J. CorsoPublished in: CoRR (2018)
Keyphrases
- weakly supervised
- video objects
- superpixels
- object segmentation
- video sequences
- object class
- relation extraction
- topic models
- object tracking
- object detectors
- object detection
- text mining
- video content
- semi supervised
- named entities
- object motion
- bounding box
- three dimensional
- object classes
- d objects
- object recognition
- computer vision
- object categories
- multiple objects
- key frames
- video frames
- segmentation method
- keywords
- multiscale