Max Robotics — Global Service Robot Directory + US Certification

About

Jun Mao is a robotics researcher whose work centers on visual navigation and place recognition for autonomous mobile robots, with a particular focus on aerial platforms. His most impactful contribution, "Learning to Fuse Multiscale Features for Visual Place Recognition" (22 citations), addresses a fundamental challenge in robotics: enabling robots to robustly recognize locations using efficient, learned visual features from convolutional neural networks. This work is critical for autonomous navigation in GPS-denied environments. Mao extends this research to aerial robots in his 2020 paper on a bio-inspired goal-directed visual navigation model (13 citations), drawing on principles from insect navigation to create lightweight, efficient systems for drones. He also tackles the broader challenge of robot adaptability in dynamic environments, proposing a reinforcement learning system that uses mixture probability and clustering distribution to improve policy learning. While his earlier work on sequence-based VPR for aerial robots (3 citations) highlights the unique difficulties of applying ground-vehicle techniques to drones—namely, the lack of a topologically ordered database—Mao’s research consistently pushes toward practical, real-world deployment of autonomous navigation systems. His work sits at the intersection of computer vision, machine learning, and bio-inspired robotics, making him a notable contributor to the field of autonomous mobile robots.