|
I'm a Ph.D student at Intelligent Robotics and Vision Lab at the University of Texas at Dallas (IRVL@UT Dallas), advised by Prof. Yu Xiang. I completed my master's degree in Electronic Engineering from University of Southern California (USC), and bachelor's degree in Electronic Information Engineering from Beijing University of Technology (BJUT). I started my research in robot learning at USC, where I spent two years worked with Prof. Joseph J. Lim, and one year with Prof. Erdem Bıyık. At USC, I worked closely with Dr.Jesse Zhang. I also interned at Horizon Robotics, working with Haonan Yu and Wei Xu. My research interests lie in robot learning and its application to enabling robots to perform tasks in everyday human life. I am particularly interested in leveraging foundation models to help robot policy learning and generalization. Google Scholar / Twitter / Linkedin / CV |
|
|
|
|
VLA-Replica: A Low-Cost, Reproducible Benchmark for Real-World Evaluation of Vision-Language-Action Models
Alex S. Huang*, Jiahui Zhang*, Shiqing Tang, Yu Xiang Preprint arXiv 2026 [Arxiv] [Website] [Paper] [Code]VLA-Replica is a low-cost, reproducible real-world benchmark for evaluating vision-language-action models. It provides an off-the-shelf hardware setup, diverse manipulation tasks, and evaluation protocols for both in-distribution and out-of-distribution testing. |
|
Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisons
Anthony Liang*, Yigit Korkmaz*, Jiahui Zhang, Minyoung Hwang, Abrar Anwar, Sidhant Kaushik, Aditya Shah, Alex S. Huang, Luke Zettlemoyer, Dieter Fox, Yu Xiang, Anqi Li, Andreea Bobu, Abhishek Gupta, Stephen Tu, Erdem Biyik, Jesse Zhang @ RSS 2026 [Arxiv] [Website] [Paper] [Code]Robometer is a general-purpose video-language dense reward model trained on RBM-1M, a dataset with over one million trajectories spanning 21 robot embodiments. It improves robot learning across online RL, offline RL, model-based RL, failure detection, and data retrieval for imitation learning. |
|
ReWiND: Language-Guided Rewards Teach Robot Policies without New Demonstrations Oral presentation @ CoRL 2025 Best Paper Award @ OOD Workshop RSS 2025 Best Paper Nomination @ RoboReps Workshop RSS 2025 [Website]We introduce an approach for learning reward functions for unseen new tasks. Our model learns a reward model based on outputs from pre-trained vision-language models, which then provides rewards for policy learning. |
|
Bootstrap Your Own Skills: Learning to Solve New Tasks with Large Language Model Guidance
Jesse Zhang,Jiahui Zhang, Karl Pertsch,Ziyi Liu,Xiang Ren,Minsuk Chang,Shao-Hua Sun,Joseph J. Lim Oral presentation (top 6.6%) @ CoRL 2023 Oral presentation @ SoCal Robotics 2023 Spotlight talk @ RSS 2023 Articulate Robots Workshop [Arxiv] [Website] [Code] [OpenReview]Our approach BOSS (BOotStrapping your own Skills) learns to accomplish new tasks by performing "skill bootstrapping," where an agent with a set of primitive skills interacts with the environment to practice new skills without receiving reward feedback for tasks outside of the initial skill set. This bootstrapping phase is guided by LLMs that inform the agent of meaningful skills to chain together. Through this process, BOSS builds a wide range of complex and useful behaviors from a basic set of primitive skills. |
|
SPRINT: Scalable Semantic Policy Pretraining via Language Instruction Relabeling
Jesse Zhang,Jiahui Zhang, Karl Pertsch,Joseph J. Lim Poster @ ICRA 2024 Spotlight talk @ CoRL 2022 LangRob Workshop Spotlight talk @ CoRL 2022 Pre-training Robot Learning Workshop [Arxiv] [Website] [Code]We propose SPRINT, a scalable approach for pre-training robot policies with a rich repertoire of skills while minimizing human annotation effort. Given a dataset of robot trajectories with an initial set of task instructions for offline pre-training, SPRINT expands the pre-training task set without additional human effort via language-model-based instruction relabeling and cross-trajectory skill chaining. |
|
Cross Domain Imitation Learning via MPC (Internship Project) Jiahui Zhang, Haonan Yu, Wei Xu [Website] We introduce CDMPC, an approach for learning new skill combinations from long-horizon skill trajectories. CDMPC enables agents to chain skills from diverse source domains and integrates them with a low-level policy in the target domain. CDMPC learns to chain short-horizon skills from long-horizon trajectories across demonstrations from diverse source domains, including various skill combinations. The policy learned from CDMPC adapts to tasks from any source domain and makes the agent able to tackle new tasks that require novel skill combinations. |
|
|
-
Served as a reviewer: IROS 2024
|
|
-
Presidential scholarship, Beijing University of Technology. 2018
Outstanding Research Achievement Award, Beijing University of Technology, Fan Gongxiu Honors College. 2017