Frankie Dong Li

I was a Research Assistant from The Intelligent Machine Research Lab (IMRL) in the SKL-IOTSC at the University of Macau, supervised by Prof. Hui Kong.

Before that, I received the B. Eng. degree from the School of Computer and Information Engineering, Hefei University of Technology in June 2022. During my undergraduate years, I was a founders of the robotics team of the Hefei University of Technology and participated in RoboMaster Robotics Competition, from 2019 to 2021. I worked as a Research Assistant for one year from June 2022 to June 2023, in the Autonomous and Multi-Agent Systems Research Lab (AMAS-Lab) at the Department of Electronic and Electrical Engineering of Southern University of Science and Technology (SUSTech), supervised by Prof. Zhiyun Lin.

When I was studying at Fuyang No.3 High School, I was sponsored as a Pearl Student by the Xin Hua Education Foundation (XHEF). I am extremely grateful to them!

Email  /  GitHub  /  Google Scholar  /  LinkedIn  /  CV  /  WeChat

Research

I'm broadly interested in the overlap between robotics and computer vision, specifically, my research focuses the on use of Deep learning to solve the real problems of Robotics applications. My current work centers on enabling robots to construct high-precision multimodal semantic maps efficiently. I believe this will establish a strong foundational framework for visual-language navigation tasks.

News

[Jan 2025]  The co-author work GARAD-SLAM has been accepted by ICRA 2025.

[Nov 2024]  The star count of our awesome-Implicit-NeRF-SLAM repository has reached 1K！

[Oct 2024]  I was invited to be a reviewer for ICRA 2025 and thank the two AEs for the assignment.

[Jun 2024]  UMAD has been accepted by IROS 2024! Thanks to everyone who participated in this project!

Selected Publications

TWC-SLAM: Multi-Agent Cooperative SLAM with Text Semantics and WiFi Features Integration for Similar Indoor Environments

TWC-SLAM is a multi-agent cooperative SLAM framework that integrates text semantics and WiFi signal features to enhance location identification and loop closure detection.

STG-Avatar: Animatable Human Avatars via Spacetime Gaussian

STG-Avatar is a 3DGS-based framework for high-fidelity animatable human avatar reconstruction.

DSOSplat: Monocular 3D Gaussian SLAM with Direct Tracking

DSOSplat is a monocular SLAM framework based on 3D Gaussian Splatting.

An Open-Source Platform for Autonomous Navigation of 4WS-4WD Agricultural Robots with Omni-Directional Kinematics

An open-source 4WS-4WD agricultural autonomous navigation robot platform that enhances maneuverability and terrain adaptability using an omni-directional kinematic model. The system integrates GNSS RTK, IMU, and steering angle encoders for high-precision localization and path tracking, validated through simulations and real-world tests. All open-source resources are available in the designated repository.

GARAD-SLAM: 3D GAussian splatting for Real-time Anti Dynamic SLAM

GARAD-SLAM is the first 3DGS-based SLAM system designed specifically for dynamic scenes.

UMAD: University of Macau Anomaly Detection Benchmark Dataset

A large-scale reference-based anomaly detection benchmark dataset that captures real-world scenarios. UMAD is applicable for both Change Detection and Anomaly Detection tasks.

Projects

Robotic Exploration for High-precision Multimodal Semantic Map

Autonomous Inspection of Underground Environments by Legged Robot

Testing the Unitree GO2 legged robot in underground environments.

GitHub Repository: awesome-Implicit-NeRF-SLAM

A comprehensive list of Implicit Representations, NeRF and 3D Gaussian Splatting papers relating to SLAM/Robotics domain, including papers, videos, codes, and related websites.

Undergraduate Project, 2018-2022

During my undergraduate studies, I worked on several projects related to robotics and electronics, including Unmanned Surface Vehicle (USV), Underwater Robot, Intelligent Sorting Trash Can, Automatic Tracking, Identification, and Measuring Device, SLAM-based Mobile Robot, and Robomaster Robots. For more information, please refer to the document.

The Applications of Fiducial Markers in Robotic Grasping: A Comparison Between AprilTag and ArUco Markers, 2023

This paper investigates the possible applications of two fiducial markers (AprilTag and ArUco) on hand eye calibration and localization procedures in a robotic grasping task and compares their performance on each procedure under different lighting conditions.

Robotics Projects at SUSTech, 2022-2023

While I was a Research Assistant at SUSTech, I developed various robotic platforms, including mobile robots and robotic arms. I utilized a range of sensors such as 2D/3D Lidar, depth cameras, millimeter-wave radar, and RFID, I implemented V-SLAM(ORB-SLAM, VINS, and RTAPMAP), Lidar-SLAM(Cartographer, LOAM/Lego-LOAM). I also developed a grasping system based on tags.

Honors & Awards

• Third Prize | International Underwater Robot Competition, 2020
• First Prize | National Undergraduate Electronics Design Contest in Anhui Province, 2020
• Second Prize | RoboMaster Robotics Competition, 2020
• Third Prize | National University Student Social Practice and Science Contest on Energy Saving & Emission Reduction, 2020
• Individual Scholarship | Hefei University of Technology, 2019, 2020, 2021

Services

Conference Reviewers

• IEEE International Conference on Robotics and Automation (ICRA) 2025 * 2
• IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2025 * 1

Journal Reviewers

• IEEE Robotics and Automation Practice * 1
• Scientific Reports * 1

Design and source code from Leonid Keselman's Jekyll fork of Jon Barron's website