Multiplane NeRF-Supervised Disentanglement of Depth and Camera Pose from Videos.
Yang FuIshan MisraXiaolong WangPublished in: CoRR (2022)
Keyphrases
- feature selection
- camera pose
- camera pose estimation
- structure from motion
- camera motion
- relative pose
- position and orientation
- feature points
- point correspondences
- urban environments
- focal length
- pose estimation
- video sequences
- depth map
- camera calibration
- bundle adjustment
- machine learning
- reference images
- line segments
- single view
- high quality
- video frames
- depth information
- camera parameters
- feature extraction
- learning algorithm
- closed form solutions
- depth images
- dynamic scenes
- d scene
- video data
- multi view
- face images
- motion estimation
- objective function