Do As I Do: Transferring Human Motion and Appearance between Monocular Videos with Spatial and Temporal Constraints.
Thiago L. GomesRenato MartinsJoão P. K. FerreiraErickson R. NascimentoPublished in: CoRR (2020)
Keyphrases
- human motion
- view invariant
- video sequences
- image sequences
- articulated human motion
- human actions
- motion capture
- motion capture data
- human motion analysis
- human body
- human pose
- body parts
- body pose
- human motion tracking
- static images
- dynamic textures
- motion sequences
- optical motion capture
- human body parts
- markerless
- motion history images
- spatio temporal
- moving camera
- motion recognition
- action recognition
- visual data
- appearance model
- temporal segmentation
- motion trajectories
- video frames
- gait recognition
- object tracking
- pose estimation
- motion analysis
- pose tracking
- human movements
- humanoid robot
- action classification
- three dimensional
- video data
- motion model
- transfer learning
- optical flow
- sports video
- visual tracking
- camera motion