Minjie Bi

Southern University of Science and Technology

Papers

Total Citations

H-Index

About

Minjie Bi is a researcher advancing the frontier of robotic perception and autonomous navigation, with a primary focus on visual simultaneous localization and mapping (SLAM) and loop closure detection (LCD). Their work addresses a critical challenge in robotics: correcting drift and accumulated errors in visual odometry, which is essential for the reliable operation of sweeping robots, drones, and autonomous vehicles. Bi’s most influential contribution is the introduction of **TLCD (Transformer-based Loop Closure Detection)**, a novel approach that leverages transformer architectures to enhance the accuracy and robustness of loop closure in visual SLAM systems. Building on this, Bi developed **TT-LCD (Tensorized-Transformer based Loop Closure Detection)**, a computationally efficient variant optimized for edge devices, enabling real-time performance in resource-constrained environments. This work is pivotal for deploying advanced SLAM in practical applications like autonomous driving and the metaverse. With over 5 citations across their top papers, Bi’s research is gaining traction in the robotics and computer vision communities. Their innovative use of transformers and tensorized models marks a significant step toward more reliable, efficient, and deployable robotic navigation systems.