Dongxu Bai

Wuhan University of Science and Technology

Papers

Total Citations

H-Index

About

Dongxu Bai is a researcher advancing the frontier of intelligent robotics, with a primary focus on simultaneous localization and mapping (SLAM) and deep learning-driven computer vision. His most cited work, “Multi-Objective Location and Mapping Based on Deep Learning and Visual SLAM” (2022, 32 citations), tackles a critical bottleneck in autonomous navigation: the poor readability and interactivity of maps built in unknown environments. By integrating deep learning with visual SLAM, Bai’s approach enables robots to not only locate themselves and construct maps but also to prioritize and distinguish primary from secondary spatial information—a leap toward truly intelligent, context-aware navigation. More recently, Bai has pushed into instance segmentation for indoor mobile robots. In his 2024 paper, “An end-to-end instance segmentation method based on improved ConvNeXt V2,” he proposes a novel RTMDet-based framework that replaces conventional backbones with the more powerful ConvNeXt V2, significantly boosting the efficiency and accuracy of real-time environmental perception. This work underscores his commitment to making robots not just autonomous, but perceptually intelligent—able to parse complex indoor scenes with human-like precision. Bai’s contributions are shaping the next generation of robotic systems that can understand and interact with their surroundings more naturally.