Yue Lin

Hi, I'm Yue Lin, this website is a brief introduction about me.
I graduated from Dalian University of Technology in 2023 with a bachelor's degree in engineering and was subsequently recommended to pursue a master's degree as a graduate student in the IIAU Lab.
My supervisor is Professor Huchuan Lu, and my research direction is robot autonomous motion planning and tracking.

Papers

GFM-Planner: Perception-Aware Trajectory Planning with Geometric Feature Metric
IROS 2025
Yue Lin, Xiaoxuan Zhang, Yang Liu, Dong Wang, Huchuan Lu
Paper | Video | Code
We propose a perception-aware trajectory planning framework with geometric feature metrics to improve the LiDAR localization accuracy of autonomous robots during navigation by enabling the robot to actively avoid areas that cause high localization errors.
Safety-First Tracker: A Trajectory Planning Framework for Omnidirectional Robot Tracking
IROS 2024
Yue Lin, Yang Liu, Pingping Zhang, Xin Chen, Dong Wang, Huchuan Lu
Paper | Video | Code
We propose a two-stage trajectory planning framework that prioritizes the robot's trajectory safety and then plans the robot's orientation to ensure target visibility. Our method enables the robot to safely follow a target in complex dynamic environments.
Eva-Tracker: ESDF-update-free, Visibility-aware Planning with Target Reacquisition for Robust Aerial Tracking
ICRA 2026 Under Review
Yue Lin, Yang Liu, Dong Wang, Huchuan Lu
Paper | Video | Code
We propose an ESDF-update-free, visibility-aware trajectory planning method for UAV target tracking, which enables UAVs to stably track targets in complex environments, avoid collisions and target occlusion, and recover lost targets effectively.

Projects

Learning Active Perception and Adversarial Game Modeling of Single-agent
NSFC Major Project
Video | Code
We use the RoboMaster2020 standard AI robot to conduct 1v1 fully automatic confrontation games. The robot can actively perceive the environment and enemy target, and establish intelligent strategies to automatically conduct confrontation games.
Multi-agent Confrontation Games in Open Environments
Tsinghua University Exchange & NSFC Major Project
Video
We developed multi-robot collaborative capture algorithms in open environments, including environmental perception, target recognition, motion planning, and strategy learning, so that the robot swarm can quickly and accurately capture an intelligent escaping target in unknown environments.
Transformer-based Lightweight Single Object Tracking
201 Institute Cooperation Project
Video | Code
We developed a high-precision, lightweight single object tracking model based on Transformer for the 201 Institute. The model shows robust performance for small targets and frequently occluded targets, and can run in real time at 110 FPS on edge devices such as JetSon Orin NX.

Competition

ICRA 2022 Sim2Real Challenge
4th in the World
Shiyao Li, Yue Lin, Zuyao You, Jiansong Pei, Zeyun Wang, Feilong Wang
Video | News
We developed lightweight visual recognition and navigation algorithms in a simulation environment, which enabled the robot to autonomously search and grab the corresponding ores, and transport them to the designated location through autonomously navigation. The algorithms developed in the simulation environment can be directly deployed on the real robot.