Detaching and Boosting: Dual Engine for Scale-Invariant Self-Supervised Monocular Depth Estimation.
Peizhe JiangWei YangXiaoqing YeXiao TanMeng WuPublished in: CoRR (2022)
Keyphrases
- scale invariant
- depth estimation
- stereo vision
- depth map
- keypoints
- stereo matching
- image sequences
- scale space
- natural images
- affine invariant
- scene understanding
- real scenes
- depth information
- dynamic scenes
- multiscale
- stereo pair
- super resolution
- pose estimation
- feature matching
- image structure
- texture images
- local binary pattern
- image matching
- invariant features
- feature selection
- high quality
- vision system
- optical flow
- stereo images
- d scene
- disparity map
- pattern recognition
- object recognition
- single image
- multi view