Adv-Depth: Self-Supervised Monocular Depth Estimation With an Adversarial Loss.
Kunhong LiZhiheng FuHanyun WangZonghao ChenYulan GuoPublished in: IEEE Signal Process. Lett. (2021)
Keyphrases
- depth estimation
- monocular images
- depth cues
- depth map
- stereo vision
- depth information
- depth perception
- stereo matching
- depth estimates
- stereo camera
- real scenes
- stereo pair
- scene understanding
- image sequences
- multi view
- dynamic scenes
- dense depth maps
- super resolution
- feature matching
- pose estimation
- stereo images
- object boundaries
- disparity map
- d scene
- depth from defocus
- high quality
- three dimensional
- motion parallax
- viewpoint
- object recognition
- computer vision
- surface reconstruction
- free viewpoint
- depth data
- ego motion
- vision system
- depth images
- belief propagation