Adversarial Training of Self-supervised Monocular Depth Estimation against Physical-World Attacks.
Zhiyuan ChengJames LiangGuanhong TaoDongfang LiuXiangyu ZhangPublished in: ICLR (2023)
Keyphrases
- physical world
- depth estimation
- monocular images
- depth map
- depth cues
- stereo vision
- virtual world
- stereo matching
- ubiquitous computing
- scene understanding
- image sequences
- depth estimates
- depth information
- dynamic scenes
- real scenes
- multi view
- pose estimation
- feature matching
- stereo pair
- human beings
- super resolution
- depth from defocus
- dynamic programming