SemanticBoost: Elevating Motion Generation with Augmented Textual Cues.
Xin HeShaoli HuangXiaohang ZhanChao WenYing ShanPublished in: CoRR (2023)
Keyphrases
- visual cues
- motion estimation
- motion cues
- image sequences
- space time
- multiple visual cues
- motion segmentation
- motion analysis
- motion detection
- motion model
- reference frame
- metadata
- mid level
- multiple cues
- motion parameters
- spatial and temporal
- camera motion
- moving objects
- high level
- medical images
- optical flow
- motion capture
- visual perception
- image motion
- human generated
- moving foreground
- motion planning
- visual content
- motion patterns
- low level
- computational complexity
- natural language
- multimedia