Ruopeng Huang

I am a Master's student in Computer Science (Intelligent Robotics) at the University of Southern California. I am currently a researcher in Prof. Erdem Biyik's LiraLab at USC. I am also a visiting student at the TASL at UC Riverside, working with Prof. Jiachen Li. And I previously did research at USC's RRoS Lab advised by Prof. S. K. Gupta.

I received my bachelor's degree in Computer Science and Technology from Beijing University of Posts and Telecommunications. My research interests include robotic manipulation, robot perception, and robot learning, with an emphasis on building robust robotic systems for real-world deployment.

🚀 I am currently applying for PhD programs for Fall 2026.

Email  /  CV  /  Scholar  /  Linkedin  

profile photo

Research
SyncTwin: Fast Digital Twin Construction and Synchronization for Safe Robotic Grasping
Ruopeng Huang, Boyu Yang, Wenlong Gui, Jeremy Morgan, Erdem Biyik, Jiachen Li
CVPR, 2026   (Under Review)
paper / video

We present SyncTwin, a fast digital twin construction and synchronization framework for safe robotic grasping in dynamic and partially observed environments.

RegraspGen: Learning General Pick and Place via Synthetic Regrasp Demonstration Generation
Mingxuan Yan*, Ruopeng Huang*,Litian Gong, Zechun Liu, Jiachen Li
RSS, 2026   (In Preparation)

RegraspGen introduces a synthetic regrasp demonstration generation pipeline that enables generalizable pick-and-place policies across diverse object geometries and task settings.

Robotic Compliant Object Prying Using Diffusion Policy Guided by Vision and Force Observations
Jeon Ho Kang, Sagar Joshi, Ruopeng Huang, Satyandra K. Gupta
IEEE Robotics and Automation Letters (RA-L), 2025   (Accepted)
paper / project page / code

This work presents a diffusion-policy-based approach for compliant object prying, integrating vision and force observations for robust contact-rich manipulation.

Energy-Aware Planning for Legged Robots Performing Logistics Tasks in Agriculture Applications
Shengqiang Chen, Yiyu Chen, Ruopeng Huang, Zishen Wei, Quan Nguyen, Satyandra K. Gupta
IEEE Transactions on Automation Science and Engineering (T-ASE), 2025   (Under Review)

We study energy-aware motion planning for legged robots in agricultural logistics scenarios, balancing task efficiency and energy consumption in wild environments.

A Systematic Study of Model Extraction Attacks on Graph Foundation Models
Haoyan Xu*, Ruizhi Qian*, Jiate Li, Ruopeng Huang, Yushun Dong, Minghao Lin, Hanson Yan, Zhengtao Yao, Qinghua Liu, Junhao Dong, Yue Zhao, Mengyuan Li
IEEE Symposium on Security and Privacy (S&P), 2026   (Under Review)
paper

This paper provides the first systematic investigation of model extraction attacks on graph foundation models, analyzing vulnerability patterns and defense implications.

Work Experience

Dorna Robotics Co. Ltd. — Robotics Software Engineer Intern
Los Angeles, United States   Apr. 2025–Jul. 2025

Designed an automatic calibration algorithm for a robotic arm without a camera; built a UI for obstacle configuration to simplify setup and make motion planning safer.
ByteDance Technology Co. Ltd. — Software Development Engineer
Beijing, China   Aug. 2021–Aug. 2022

Refactored the Hawk platform from Python to Go, focusing on architecture and concurrency to resolve efficiency issues for DouYin (TikTok) content security.
Baidu Technology Co. Ltd. — Development and Test Engineer
Beijing, China   Oct. 2020–Aug. 2021

Implemented algorithm evaluation in Python and maintained backend platform services in Go for the Vision Technology Center.

Course Projects

Multimodal Model-Based Robotic Arm Control in Household Scenarios
Deep Learning and Application (Instructor: Zhao Yue)   Jan. 2025–May 2025

Built an OpenVLA-based VLA to control a robotic arm with natural language; evaluated using the LIBERO benchmark suite.
The Application of AI Algorithms in Robotic Arms
Foundation of Artificial Intelligence (Instructor: Raj Venkat)   Aug. 2023–Dec. 2023

Implemented motion planning for a 3-axis robotic arm in 2D using search, MDP, Q-learning, DDPG, and related RL/control methods.
The Application of Multi-Sensor Fusion in Mobile Robotics
Mobile Robotics (Instructor: David Rosen)   Aug. 2023–Dec. 2023

Augmented LiDAR SLAM with GNSS factors in a fusion pipeline to improve localization accuracy and mapping quality.

Service & Skills

Service

  • IROS 2025 Reviewer.

Skills

  • Python, C/C++, ROS2, Isaac Sim, ManiSkill, MuJoCo, MATLAB, Java, VHDL.

Hardware Experience

  • Robots: KUKA iiwa, Franka Panda, Aloha Dual Arm Robot, LeRobot, LEGO Robot.

Webpage inspired by Jon Barron