PackNet-SfM: 3D Packing for Self-Supervised Monocular Depth Estimation.
Vitor GuiziliniRares AmbrusSudeep PillaiAdrien GaidonPublished in: CoRR (2019)
Keyphrases
- depth estimation
- monocular images
- depth cues
- structure from motion
- stereo vision
- depth map
- depth estimates
- image sequences
- depth perception
- stereo matching
- depth information
- dynamic scenes
- scene understanding
- missing data
- real scenes
- camera motion
- bundle adjustment
- super resolution
- stereo pair
- multi view
- computer vision
- disparity map
- ego motion
- d scene
- stereo images
- depth from defocus
- model based pose estimation
- specular reflection
- pose estimation
- high quality
- image matching
- object tracking
- optical flow
- three dimensional