Deep introspective SLAM: deep reinforcement learning based approach to avoid tracking failure in visual SLAM.
Kanwal NaveedMuhammad Latif AnjumWajahat HussainDonghwan LeePublished in: Auton. Robots (2022)
Keyphrases
- visual slam
- monocular slam
- real time
- dynamic environments
- reinforcement learning
- simultaneous localization and mapping
- monocular camera
- camera motion
- bundle adjustment
- particle filter
- dynamic programming
- visual odometry
- motion segmentation
- multibody
- single camera
- data association
- optical flow
- state space
- motion blur
- motion analysis
- visual tracking
- kalman filter
- mobile robot