Jingxiang Guo

Jingxiang Guo | 郭京翔

I am a PhD student at the National University of Singapore, in the Language and Vision Lab (LV-Lab), advised by Prof. Yan Shuicheng and co-supervised by Prof. Jianshu Zhou (G&M Lab). I received my M.Comp. in AI from NUS and my B.Eng. in Automation from HITSZ.

Email / CV / GitHub / Google Scholar / More Academic Links / WeChat

News

[2026/04] 🎉 Honored to receive PhD offers with scholarships from IC, EPFL, NUS, and NTU. I chose NUS.
[2026/01] 🎉 Goal-VLA, AdaptPNP, and ShapeForce were accepted to ICRA 2026!
[2025/11] 🎉 DexSinGrasp was accepted to RA-L!
[2025/10] 🏅 SayWhen won the Best Demo Award Finalist at IROS Workshop on Contact and Impact-aware Manipulation!
[2025/06] 🎉 MetaFold was accepted to IROS 2025 as Oral Presentation!
[2025/05] 🏅 TelePreview won the Best Paper Award at ICRA 2025 Workshop on Human-Centric Multilateral Teleoperation!
[2025/05] 🏅 $\mathcal{D(R, O)}$ Grasp won the ICRA 2025 Best Paper Award on Robot Manipulation and Locomotion!
[2025/04] 🎉 Manual2Skill was accepted to RSS 2025!
[2024/12] 🎉 $\mathcal{D(R, O)}$ Grasp was accepted to ICRA 2025!
[2024/11] 🏅 $\mathcal{D(R, O)}$ Grasp won the Best Robotics Paper Award at CoRL 2024 MAPoDeL Workshop!

Research

My research interests lie in 🤖 robot multi-modal learning, 🦾 dexterous manipulation, and 🤝 human-robot interaction. My long-term goal is to create truly conscious robotic life, pushing the boundaries of what's possible with machines.

Research Proposal 🔗

Research Map

World Model Framework

Selected Publications

Papers sorted by recency. Representative papers are highlighted.

	Goal-VLA: Image-Generative VLMs as Object-Centric World Models Empowering Zero-shot Robot Manipulation Haonan Chen, Jingxiang Guo, Bangjun Wang, Tianrui Zhang, Xuchuan Huang, Boren Zheng, Yiwen Hou, Chenrui Tie, Jiajun Deng, Lin Shao Website / arXiv / Code ICRA 2026 International Conference on Robotics and Automation Poster Presentation, IROS 2025 @ Human-in-the-loop Robot Learning TL;DR: A zero-shot framework that leverages Image-Generative VLMs as world models to generate desired goal states for generalizable manipulation, using object state representation as the interface between high-level and low-level policies.
	AdaptPNP: Integrating Prehensile and Non-Prehensile Skills for Adaptive Robotic Manipulation Jinxuan Zhu, Chenrui Tie, Xinyi Cao, Yuran Wang, Jingxiang Guo, Zixuan Chen, Haonan Chen, Junting Chen, Yangyu Xiao, Ruihai Wu, Lin Shao Website / arXiv / Code ICRA 2026 International Conference on Robotics and Automation TL;DR:* A VLM-empowered framework that combines prehensile and non-prehensile skills using a digital twin for proactive manipulation planning.
	ShapeForce: Low-Cost Soft Robotic Wrist for Contact-Rich Manipulation Jinxuan Zhu, Zihao Yan, Yangyu Xiao, Jingxiang Guo, Chenrui Tie, Xinyi Cao, Yuhang Zheng, Lin Shao Website / arXiv / Code ICRA 2026 International Conference on Robotics and Automation Best Demo Award Finalist, IROS 2025 @ Contact and Impact-aware Manipulation TL;DR: A low-cost, plug-and-play soft wrist that provides force-like signals for contact-rich robotic manipulation by converting external force into measurable deformations.
	QuaMa: Extending Skill Learning with Delta Dynamics Modeling for Precise and Generalizable Quantitative Manipulation Yiwen Hou, Yaoyu Cheng, Jiayu Luo, Zhixuan Xu, Chongkai Gao, Jingxiang Guo, Lin Shao Website / arXiv / Code Under Review TL;DR:* A framework that combines behavior cloning with delta dynamics modeling for precise quantitative manipulation, achieving sub-gram accuracy in pouring tasks with robust zero-shot generalization.
	InsertScale: A Benchmark towards Foundation Visuo-Tactile Policy Scalable Learning for Insertion Task Congsheng Xu, Jingxiang Guo, Baijun Chen, Liuhaichen Yang, Zhen Zou, Yuzhang Li, Jieji Ren, Yiming Wang, Yichao Yan, Yao Mu, Xiaokang Yang Poster / arXiv / Code Late Breaking Results Poster @ IROS 2025 Oral Presentation, IROS 2025 @ Learning from Teleoperation TL;DR:* Propose a visuo-tactile benchmark and scalable policy learning framework for insertion tasks, providing standardized tasks, metrics, and multi-modal datasets to evaluate and train foundation policies.
	Benchmarking Generalizable Bimanual Manipulation: RoboTwin Dual-Arm Collaboration Challenge at CVPR 2025 MEIS Workshop Robotwin Challenge Organizers, RoboTwin Challenge Volunteers (Jingxiang Guo, etc.), RoboTwin Challenge Participants Website / arXiv / Media (量子位) Tech Report @ CVPR 2025 MEIS Workshop TL;DR: Official Technical Report of RoboTwin Dual-Arm Collaboration Challenge
	World4Omni: A Zero-Shot Framework from Image Generation World Model to Robotic Manipulation Haonan Chen, Bangjun Wang, Jingxiang Guo, Tianrui Zhang, Yiwen Hou, Xuchuan Huang, Chenrui Tie, Lin Shao Website / Early Idea / arXiv / Code Poster Presentation, ICML 2025 @ Building Physically Plausible World Models TL;DR:* Propose a novel framework that leverages a pre-trained multimodal image-generation model as a world model to guide policy learning.
	DexSinGrasp: Learning a Unified Policy for Dexterous Object Singulation and Grasping in Cluttered Environments Lixin Xu, Zixuan Liu, Zhewei Gui, Jingxiang Guo, Zeyu Jiang, Zhixuan Xu, Chongkai Gao, Lin Shao Website / arXiv / Code RA-L Robotics and Automation Letters Spotlight Presentation, ICRA 2025 @ Handy Moves: Dexterity in Multi-Fingered Hands TL;DR: Implement a unified policy for dexterous object singulation and grasping in cluttered environments, enabling robots to handle complex manipulation tasks with high success rates.
	Manual2Skill: Learning to Read Manuals and Acquire Robotic Skills for Furniture Assembly Using Vision-Language Models Chenrui Tie, Shengxiang Sun, Jinxuan Zhu, Yiwei Liu, Jingxiang Guo, Yue Hu, Haonan Chen, Junting Chen, Ruihai Wu, Lin Shao Website / arXiv / Code / Media (机器之心) RSS 2025 Robotics: Science and Systems Oral Presentation, CVPR 2025 @ 3D Vision Language Models for Robotic Manipulation TL;DR: Propose a novel approach that leverages vision-language models to interpret assembly manuals and translate them into executable robotic skills for furniture assembly tasks.
	TelePreview: A User-Friendly Teleoperation System with Virtual Arm Assistance for Enhanced Effectiveness Jingxiang Guo, Jiayu Luo, Zhenyu Wei, Yiwen Hou, Zhixuan Xu, Xiaoyi Lin, Chongkai Gao, Lin Shao Website / arXiv / Code (Coming soon) Best Paper Award, ICRA 2025 @ Human-Centric Multilateral Teleoperation Spotlight Presentation, ICRA 2025 @ Human-Centric Multilateral Teleoperation TL;DR:* Implement a low-cost teleoperation system utilizing data gloves and IMU sensors, paired with an assistant module that improves data collection process by visualizing future robot operations through visual previews.
	MetaFold: A Closed-loop Pipeline for Universal Clothing Folding via End-to-end Point Cloud Trajectory Generation Haonan Chen, Junxiao Li, Chongkai Gao, Zhixuan Xu, Chenting Wang, Yiwen Hou, Jingxiang Guo, Shensi Xu, Jiaqi Huang, Weidong Wang, Lin Shao Website / arXiv / Dataset IROS 2025 International Conference on Intelligent Robots and Systems Oral Presentation @ IROS 2025 TL;DR: Propose a closed-loop pipeline for universal clothing folding using end-to-end point cloud trajectory generation, enabling robots to handle various types of clothing with high precision.
	$\mathcal{D(R,O)}$ Grasp: A Unified Representation of Robot and Object Interaction for Cross-Embodiment Dexterous Grasping Zhenyu Wei, Zhixuan Xu, Jingxiang Guo, Yiwen Hou, Chongkai Gao, Zhehao Cai, Jiayu Luo, Lin Shao Website / arXiv / Code / Media (机器之心) ICRA 2025 International Conference on Robotics and Automation ICRA 2025 Best Paper Award on Robot Manipulation and Locomotion Best Robotics Paper Award, CoRL 2024 @ MAPoDeL Oral Presentation, CoRL 2024 @ MAPoDeL Spotlight Presentation, CoRL 2024 @ LFDM × 🔗 Verify Individual Certificate TL;DR: Introduce $\mathcal{D(R,O)}$, a novel interaction-centric representation for dexterous grasping tasks that goes beyond traditional robot-centric and object-centric approaches, enabling robust generalization across diverse robotic hands and objects.
	MASQ: Multi-Agent Reinforcement Learning for Single Quadruped Robot Locomotion Qi Liu, Jingxiang Guo, Sixu Lin, Shuaikang Ma, Jinxuan Zhu, Yanjie Li arXiv / Video / Media (具身智能之心) ICML 2025 Workshop @ NewInML TL;DR: Introduce MASQ, a novel approach using multi-agent reinforcement learning (MARL) for single quadruped robot locomotion. By treating each leg as an independent agent, MASQ accelerates learning and boosts real-world robustness, surpassing traditional methods.
	Multi-Agent Target Assignment and Path Finding for Intelligent Warehouse: A Cooperative Multi-Agent Deep Reinforcement Learning Perspective Qi Liu, Jianqi Gao, Dongjie Zhu, Zhongjian Qiao, Jingxiang Guo, Pengbin Chen, Yanjie Li arXiv / Media In submission TL;DR: Develop a cooperative multi-agent deep reinforcement learning approach for intelligent warehouse systems, focusing on efficient target assignment and path finding for multiple robots.
	Logarithmic Function Matters Policy Gradient Deep Reinforcement Learning Qi Liu, Jingxiang Guo, Zhongjian Qiao, Pengbin Chen, Yanjie Li Paper / Code DAI 2024 Distributed AI (DAI) conference Oral Presentation @ DAI 2024 TL;DR: Investigate the impact of logarithmic functions in policy gradient deep reinforcement learning, demonstrating improved performance and stability in various RL tasks.
	Momentum Prediction for Tennis Matches Based on Counter-Factual Analysis and Multi-LGBM Jingxiang Guo, Jinxuan Zhu, Sixu Lin, Feng Shi IEEE Xplore / Code ICAACE 2024 International Conference on Advanced Algorithms and Control Engineering TL;DR: Develop a novel approach for tennis match momentum prediction using counter-factual analysis and multi-LGBM models, achieving improved accuracy in match outcome predictions.
	ECAPA-TDNN Embeddings for Speaker Recognition Jingxiang Guo, Jinxuan Zhu, Sixu Lin, Feng Shi IEEE Xplore / Code AINIT 2024 International Seminar on Artificial Intelligence, Networking and Information Technology TL;DR: Implement and evaluate ECAPA-TDNN embeddings for speaker recognition tasks, demonstrating improved performance in speaker verification and identification.
	Quick reversing device and quick track reversing device Kuntian Dai, Jingxiang Guo, Nengfeng Liu, Guanyu Hou, Jinbin Guo, Junkai Wang, Ruiquan Dong Google Patents / Certificate Invention Patent National Patent CN116000896B TL;DR: Design and patent a novel quick reversing device and track reversing system, improving efficiency and safety in industrial applications.

Award

Best Demo Award Finalist, IROS 2025 @ Contact and Impact-aware Manipulation Jun 2025
Best Paper Award, ICRA 2025 @ Human-Centric Multilateral Teleoperation May 2025
ICRA 2025 Best Paper Award on Robot Manipulation and Locomotion May 2025
Best Robotics Paper Award, CoRL 2024 @ Morphology-Aware Policy and Design Learning Nov 2024
National First Prize, RoboMaster University Championship (RMUC) Aug 2023
National First Prize, China Intelligent Robots Innovation Competition Jul 2023
Honorable Mention, Mathematical Contest in Modeling (MCM) Sep 2023
Provincial First Prize, China Undergraduate Mathematical Contest in Modeling (CUMCM) Sep 2023
First Prize, RoboMaster2023 Infantry Match Mar 2023
First Prize, RoboMaster2022 (National Final) Aug 2022
First Prize, RoboMaster2022 (Central Region) Jun 2022
First Prize, RoboMaster2023 (High School League) Mar 2023
First Prize, Hongli Cup Mathematical Modelling Mar 2023
Second Prize, National Embedded Chip Design Jun 2023
Second Prize, China Robotics Competition 2023 Sep 2023
Second Prize, Mechanical Engineering Innovation 2023 Sep 2023
Second Prize, iCAN Innovation and Entrepreneurship Nov 2023
Outstanding Student of the Year 2021-2022 Nov 2022
HITSZ Academic Scholarship Nov 2023
Outstanding Member of the League 2022-2023 May 2023