CrossAdapt: Cross-Scene Adaptation for Multi-Domain Depth Estimation.
Yu ZhangMuhammad Usman RafiqueGordon A. ChristieNathan JacobsPublished in: IGARSS (2023)
Keyphrases
- depth estimation
- multi domain
- stereo vision
- depth map
- scene understanding
- dynamic scenes
- stereo matching
- real scenes
- domain specific
- cross domain
- depth information
- stereo pair
- super resolution
- d scene
- object detection
- object recognition
- multi view
- depth from defocus
- dynamic programming
- feature matching
- heterogeneous networks
- computer vision
- space time
- post processing
- vision system
- three dimensional
- disparity map
- video sequences
- viewpoint
- moving objects
- high quality
- machine learning
- depth estimates