SelfVIO: Self-supervised deep monocular Visual-Inertial Odometry and depth estimation.
Yasin AlmaliogluMehmet TuranMuhamad Risqi Utama SaputraPedro Porto Buarque de GusmãoAndrew MarkhamNiki TrigoniPublished in: Neural Networks (2022)
Keyphrases
- depth estimation
- monocular images
- inertial sensors
- stereo vision
- depth cues
- depth map
- depth perception
- stereo matching
- depth information
- pose estimation
- depth estimates
- position and orientation
- stereo pair
- image sequences
- scene understanding
- stereo camera
- dynamic scenes
- real scenes
- visual odometry
- super resolution
- multi view
- disparity map
- feature matching
- object recognition
- three dimensional
- stereo images
- human body
- real time
- computer vision
- depth from defocus
- high quality
- video sequences
- feature points
- object tracking