Unsupervised Learning of Monocular Depth Estimation with Bundle Adjustment, Super-Resolution and Clip Loss.
Lipu ZhouJiamin YeMontiel AbelloShengze WangMichael KaessPublished in: CoRR (2018)
Keyphrases
- super resolution
- depth estimation
- bundle adjustment
- unsupervised learning
- structure from motion
- image sequences
- depth map
- camera motion
- low resolution
- multi view
- high resolution
- least squares
- motion estimation
- camera parameters
- point correspondences
- pose estimation
- semi supervised
- image restoration
- high quality
- single image
- motion blur
- optical flow
- light field
- machine learning
- d scene
- expectation maximization
- dynamic scenes
- object recognition
- dimensionality reduction
- maximum likelihood
- video sequences
- depth information
- computer vision
- multiple views
- camera calibration
- viewpoint
- missing data
- point cloud
- face recognition
- feature selection
- feature points