Multi-Task Learning based on Separable Formulation of Depth Estimation and its Uncertainty.
Akari AsaiDaiki IkamiKiyoharu AizawaPublished in: CVPR Workshops (2019)
Keyphrases
- multi task learning
- depth estimation
- multi task
- stereo vision
- depth map
- stereo matching
- learning tasks
- gaussian processes
- depth information
- learning problems
- real scenes
- high order
- scene understanding
- dynamic scenes
- transfer learning
- theoretical analysis
- super resolution
- learning algorithm
- multi view
- higher order
- learning models
- stereo pair
- data sets
- supervised learning
- probabilistic model
- reinforcement learning
- high quality
- decision trees
- machine learning