ROFT: Real-Time Optical Flow-Aided 6D Object Pose and Velocity Tracking.
Nicola A. PigaYuriy OnyshchukGiulia PasqualeUgo PattaciniLorenzo NatalePublished in: CoRR (2021)
Keyphrases
- d objects
- optical flow
- pose estimation
- physically plausible
- motion model
- markerless
- motion segmentation
- viewpoint
- pose parameters
- multi view
- object recognition
- image sequences
- velocity field
- d mesh
- range data
- three dimensional
- human pose
- pose normalization
- camera motion
- moving camera
- multiple views
- computer vision
- object tracking
- cad models
- particle filter
- camera pose
- flow field
- motion analysis
- d scene
- motion field
- ego motion
- moving objects
- vector field
- object classes
- motion parameters
- surface patches
- object representation
- topological information
- degrees of freedom
- integral imaging
- appearance model