Monocular Depth Estimation with a Multi-task and Multiple-input Architecture Using Depth Gradient.
Michiru TakamineSatoshi EndoPublished in: SCIS/ISIS (2020)
Keyphrases
- depth estimation
- multi task
- multiple input
- depth cues
- depth map
- stereo vision
- depth information
- stereo matching
- depth estimates
- learning tasks
- scene understanding
- image sequences
- dynamic scenes
- super resolution
- stereo pair
- multi view
- pose estimation
- real scenes
- transfer learning
- multi class
- feature matching
- learning problems
- feature selection
- stereo images
- disparity map
- d scene
- specular reflection
- three dimensional
- depth from defocus
- object boundaries
- active learning
- optical flow
- low resolution
- machine learning