Publications by authors named "HyungGi Jo"
Uncertainty-Aware Depth Network for Visual Inertial Odometry of Mobile Robots.
Sensors (Basel)
October 2024
Article Synopsis
- Simultaneous localization and mapping (SLAM) is essential for autonomous vehicles and mobile robots, often utilizing a combination of sensors, including inertial measurement units (IMUs) for motion estimation.
- The paper introduces an uncertainty-aware depth network (UD-Net) that estimates both depth and uncertainty maps, enhancing visual-inertial odometry (VIO) performance by filtering out unreliable depth values.
- Experiments with UD-Net on the KITTI dataset and a custom dataset showed substantial improvements over traditional VIO methods, confirming its effectiveness in providing accurate mapping and localization.