SDF-SLAM: A Deep Learning Based Highly Accurate SLAM Using Monocular Camera Aiming at Indoor Map Reconstruction With Semantic and Depth Fusion.
Chen YangQi ChenYaoyao YangJingyu ZhangMinshun WuKuizhi MeiPublished in: IEEE Access (2022)
Keyphrases
- highly accurate
- visual slam
- monocular camera
- deep learning
- visual odometry
- simultaneous localization and mapping
- indoor environments
- real time
- dynamic environments
- mobile robot
- depth images
- ground plane
- ego motion
- bundle adjustment
- high quality
- high accuracy
- unsupervised learning
- inertial sensors
- autonomous navigation
- feature tracking
- camera motion
- machine learning
- kalman filter
- long range
- depth map
- object recognition
- three dimensional
- data association
- depth information
- structure from motion
- mobile robotics
- multibody
- feature selection
- computer vision
- weakly supervised
- state space
- high resolution
- image sequences