Subspace-PnP: A Geometric Constraint Loss for Mutual Assistance of Depth and Optical Flow Estimation.
Cheng ChiTianyu HaoQingjie WangPeng GuoXin YangPublished in: Int. J. Comput. Vis. (2022)
Keyphrases
- optical flow estimation
- geometric constraints
- feature matching
- optical flow
- flow field
- epipolar geometry
- coarse to fine
- motion field
- motion model
- stereo matching
- multiple images
- multiscale
- camera calibration
- principal component analysis
- velocity field
- depth map
- motion estimation
- ground plane
- vector field
- object recognition
- image sequences
- computer vision
- motion analysis
- image pairs
- feature extraction
- face recognition