InMoDeGAN: Interpretable Motion Decomposition Generative Adversarial Network for Video Generation.
Yaohui WangFrançois BrémondAntitza DantchevaPublished in: CoRR (2021)
Keyphrases
- space time
- spatial and temporal
- motion analysis
- object motion
- image sequences
- input video
- key frames
- shot change detection
- real time
- temporal filtering
- moving camera
- motion features
- multimedia
- video sequences
- camera motion
- video frames
- spatio temporal
- motion trajectories
- video data
- dynamic scenes
- image frames
- video content
- wireless sensor networks
- temporal consistency
- video streams
- motion model
- video summarization
- motion detection
- motion patterns
- video clips
- video retrieval
- visual cues
- moving objects
- motion segmentation
- video surveillance
- optical flow
- single frame
- temporal coherence
- network traffic
- multimedia data
- video footage
- peer to peer
- computer vision