ProDepth: Boosting Self-Supervised Multi-Frame Monocular Depth with Probabilistic Fusion.
Sungmin WooWonjoon LeeWoo Jin KimDogyoon LeeSangyoun LeePublished in: CoRR (2024)
Keyphrases
- multi frame
- image sequences
- optical flow
- motion estimation
- super resolution
- depth map
- point correspondences
- depth cues
- multiple frames
- motion analysis
- depth information
- ego motion
- computationally efficient
- single frame
- stereo correspondence
- optic flow
- motion parameters
- filtering algorithm
- closed form
- feature points
- high resolution
- depth estimation
- machine learning
- planar scenes
- motion detection
- flow field
- motion segmentation
- video frames
- low resolution
- three dimensional
- image processing
- computer vision