Adversarial Training of Self-supervised Monocular Depth Estimation against Physical-World Attacks.
Zhiyuan ChengJames LiangGuanhong TaoDongfang LiuXiangyu ZhangPublished in: CoRR (2023)
Keyphrases
- physical world
- depth estimation
- monocular images
- depth cues
- depth map
- stereo vision
- virtual world
- ubiquitous computing
- stereo matching
- depth information
- depth estimates
- pose estimation
- stereo pair
- image sequences
- real scenes
- dynamic scenes
- scene understanding
- feature matching
- information processing
- space time
- stereo images
- specular reflection
- multi view
- input data
- optical flow
- video sequences