A self-supervised monocular odometry with visual-inertial and depth representations.
Lingzhe ZhaoTianyu XiangZhuping WangPublished in: J. Frankl. Inst. (2024)
Keyphrases
- inertial sensors
- depth cues
- visual odometry
- sensor fusion
- visual cues
- visual representations
- low level
- visual features
- visual information
- mid level
- pose estimation
- dynamic model
- ego motion
- relative depth
- model based tracking
- depth perception
- motion tracking
- image sequences
- position and orientation
- depth information
- depth map
- optical flow
- autonomous navigation
- depth images
- motion sequences
- depth ordering
- motion estimation
- object recognition