Recovering 3D Human Poses and Camera Motions from Deep Sequence.
Takashi ShimizuFumihiko SakaueJun SatoPublished in: VISIGRAPP (5: VISAPP) (2018)
Keyphrases
- camera motion
- human pose
- egomotion estimation
- image sequences
- pose estimation
- human body
- action recognition
- video sequences
- epipolar geometry
- body parts
- ego motion
- moving objects
- optical flow
- motion parameters
- feature points
- depth images
- human detection
- random fields
- structure from motion
- human motion
- human actions
- graph structure