Robotics · Mapping · Perception · Navigation
PhD Candidate · Wuhan University
My research interests lie in embodied AI navigation, with a focus on visual semantic SLAM, semantic perception, vision-language navigation, static and dynamic object perception, and spatial intelligence. I have independently developed several semantic localization and navigation systems. I am currently pursuing a PhD at Wuhan University under the supervision of Professor Chi Guo at the School of Robotics. I expect to graduate in late 2026 and am currently seeking career opportunities. Please feel free to reach out to connect or exchange ideas.
Research systems & demonstrations
A static and dynamic object-level active visual SLAM system integrating interactive navigation. It maintains geometric and semantic scene memory and combines LLM/VLM reasoning for object-goal navigation and instruction following.
The demonstrations cover mapping, language-guided navigation, dynamic target following, and validation.
Visual SLAM fusing static and dynamic semantic information for driving scenes. The demonstrations show road-aware reconstruction alongside KITTI odometry and tracking sequences.
The project connects scene semantics with geometric estimation in dynamic traffic environments.
A mesh-based dynamic visual SLAM and multi-object tracking method. Mesh structure helps represent and follow moving objects while the camera estimates its own motion.
The demos show mesh generation, system running on an OMD sequence, and four KITTI tracking sequences.
Camera localization combined with tracking of multiple moving objects. The demos show both the lightweight version and the full version with local bundle adjustment (BA).
This project organizes a scene around recognizable objects instead of only sparse geometric features. Indoor and outdoor demos show object-aware mapping in different settings.
A compact odometry system combining camera and inertial measurements to estimate motion. The MH01 sequence illustrates its trajectory and map output.