Motion Guided Region Message Passing for Video Captioning.
Shaoxiang ChenYu-Gang JiangPublished in: ICCV (2021)
Keyphrases
- message passing
- space time
- belief propagation
- key frames
- distributed systems
- motion estimation
- video sequences
- shared memory
- optical flow
- video content
- image sequences
- markov random field
- probabilistic inference
- factor graphs
- video frames
- human motion
- sum product
- motion model
- graphical models
- video streams
- video data
- approximate inference
- distributed shared memory
- inference in graphical models
- graph cuts
- motion vectors
- input image
- pairwise
- object recognition