Self-supervised Depth Estimation Leveraging Global Perception and Geometric Smoothness Using On-board Videos.
Shaocheng JiaXin PeiWei YaoS. C. WongPublished in: CoRR (2021)
Keyphrases
- depth estimation
- dynamic scenes
- stereo vision
- depth map
- depth cues
- stereo matching
- video sequences
- scene understanding
- real scenes
- feature matching
- depth information
- model based pose estimation
- stereo pair
- high quality
- background subtraction
- video surveillance
- multi view
- geometric constraints
- image pairs
- stereo images
- three dimensional
- video frames
- video data
- post processing
- image sequences