Self-supervised depth estimation with uncertainty-weight joint loss function based on laparoscopic videos.
Wenda LiYuichiro HayashiMasahiro OdaTakayuki KitasakaKazunari MisawaKensaku MoriPublished in: Medical Imaging: Image-Guided Procedures (2022)
Keyphrases
- loss function
- depth estimation
- dynamic scenes
- stereo vision
- depth map
- pairwise
- stereo matching
- depth information
- scene understanding
- video sequences
- support vector
- real scenes
- multi view
- video frames
- stereo pair
- feature matching
- video surveillance
- space time
- video data
- super resolution
- image sequences
- object recognition
- motion segmentation
- image restoration
- moving objects