Papers
6
Total Citations
58
H-Index
4
About
Jun Cheng is a robotics and computer vision researcher whose work centers on simultaneous localization and mapping (SLAM), sensor fusion, and autonomous navigation systems. His most recognized contribution, "LMVI-SLAM: Robust Low-Light Monocular Visual-Inertial SLAM" (2019, 19 citations), addresses one of the field's persistent challenges — maintaining reliable localization in degraded lighting conditions — by leveraging the complementary strengths of visual and inertial sensors. This work, alongside his improved initialization method for monocular VI-SLAM (2021, 12 citations), demonstrates a sustained commitment to pushing the boundaries of visual-inertial odometry in real-world, demanding environments. Cheng has also made practical contributions to 3D lidar sensing, developing a low-cost calibration technique for rotating 2D lidar systems (2021, 18 citations) that reduces point cloud error without requiring real-time motor shaft angle measurement — a meaningful advancement for cost-sensitive robotics applications. His earlier work in semantic mapping and RGB-D SLAM reflects a broader interest in building perception systems that combine geometric accuracy with scene understanding. Collectively, Cheng's research positions him as a versatile contributor to mobile robotics perception, with growing influence across both academic and applied autonomous systems communities.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3An Improved Initialization Method for Monocular Visual-Inertial SLAM12 citations · 2021
- 4
- 5A Robust RGB-D Image-Based SLAM System3 citations · 2017
- 6A Lifted Semi-Direct Monocular Visual Odometry2 citations · 2019