Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions.
Shuang LiYilun DuAntonio TorralbaJosef SivicBryan C. RussellPublished in: CoRR (2021)
Keyphrases
- weakly supervised
- object detectors
- object detection
- space time
- spatial and temporal
- object class
- human actions
- topic models
- superpixels
- event detection
- video sequences
- video frames
- relation extraction
- video data
- input image
- semi supervised
- spatio temporal
- object categories
- region of interest
- bounding box
- object segmentation
- low level
- event recognition