Hi, I'm Tianxing CHEN (陈天行) 👋
<!-- -->

🏠 Homepage: tianxingchen.github.io

Hi! I am Tianxing Chen (Chinese name: 陈天行), Co-Founder and CTO of Xspark AI, founder of
Lumina Embodied AI Community,
and also a first-year Ph.D. student at
MMLab@HKU,
The University of Hong Kong (since Oct. 2025), supervised by
Prof. Ping Luo.
My research focuses on Embodied AI Infra, including Robotic Foundation Model, Data Generator and Evaluation.

I have published multiple papers at top venues such as CVPR, RSS, ICLR, ICRA and ACL.
My works received Best Paper Award at the ECCV 2024 MAAS Workshop,
Champion at NeurIPS 2025 MARS Challenge (Control Track), third-runnerup at IROS 2025 AgiBot World Challenge (Manipulation Track),
Best Poster Award at IROS 2025 RoMaDo, RoDGE, AIR4S Workshop and ChinaSI 2025,
and a CVPR Highlight recognition.
I organized Embodied-AI-Guide project, and received over 10k Github stars,
which is the most popular EAI tutorial.
Beyond research, I have been honored with the CCF Elite Collegiate Award (99 nationwide)
and the "Excellent Star of SZU", the highest undergraduate honor at Shenzhen University (ranked 2/30,000).
I also won a silver medal in the ACM-ICPC regional competition and was national runner-up in the
RAICOM robotics developer competition.

I have served as an organizer of the SIGGRAPH ASIA 2025 TriFusion Workshop, CVPR 2025 MEIS Workshop,
the CVPR 2025 RoboTwin Dual-Arm Collaboration Challenge, and the Greater Bay Area Embodied AI Workshop.
I am also a student committee member of the Autonomous Robotic Technology Seminar (ARTS).
In addition, I share my thoughts on research and technology as a blogger and vlogger,
with over 20k followers on Rednote 🎒@TianxingChen (陈天行).

Chinese Profile (中文个人介绍)
Hi！我是陈天行 (English name: Tianxing Chen)，Xspark AI 联合创始人兼CTO，Lumina具身智能社区的创始人，
同时也是香港大学MMLab@HKU的一年级博士生（始于2025年9月），
师从罗平教授。
我的研究方向是具身智能基础架构，包括机器人通用基础模型、数据合成器与评测。
我在CVPR、RSS、ICLR、ICRA、ACL等学术顶会发表多篇论文。在ECCV 2024 MAAS Workshop上获得最佳论文奖，
获得NeurIPS 2025 MARS操作赛冠军，IROS 2025 AgiBot World操作赛季军，在IROS 2025 RoMaDo、RoDGE、AIR4S Workshop
与ChinaSI 2025上获得最佳海报奖，获得CVPR Highlight等。我组织了Embodied-AI-Guide项目，获得了超过一万Github Stars，
是最流行的具身智能教程。我曾经获得CCF优秀大学生（全国年度99人）、荔园卓越之星（深圳大学本科生最高荣誉）等荣誉。
竞赛上获得ACM-ICPC国际大学生程序设计竞赛银牌，在RAICOM机器人开发者大赛中获得赛道全国亚军等。
我也曾作为组织者举办SIGGRAPH ASIA 2025 TriFusion Workshop、CVPR 2025 MEIS Workshop、
CVPR 2025 RoboTwin双臂协作竞赛、大湾区具身智能研讨会等。担任ARTS自主机器人技术研讨会的学生委员会成员。
同时，我还在小红书拥有超20k粉丝🎒@TianxingChen (陈天行)。

Feel free to contact me by email if you are interested in discussing or collaborating with me.

&nbsp;
&nbsp;
&nbsp;
&nbsp;
&nbsp;
&nbsp;
&nbsp;

📑 Papers

First / Co-first Author

• [arXiv 26] ManiTwin: Scaling Data-Generation-Ready Digital Object Dataset to 100K, website
• [arXiv 26] RMBench: Memory-Dependent Robotic Manipulation Benchmark with Insights into Policy Design, website
• [arXiv 26] UniVTAC: A Unified Simulation Platform for Visuo-Tactile Manipulation Data Generation, Learning, and Benchmarking, website
• [arXiv 25] FieldGen: From Teleoperated Pre-Manipulation Trajectories to Field-Guided Data Generation, website
• [ICML 26] RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation, Webpage / Code
• [Technical Report] Benchmarking Generalizable Bimanual Manipulation: RoboTwin Dual-Arm Collaboration Challenge at CVPR 2025 MEIS Workshop, Webpage
• [CVPR 25 (Highlight)] RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins, Webpage / Code
• [CVPR 25] G3Flow: Generative 3D Semantic Flow for Pose-aware and Generalizable Object Manipulation, Webpage, Code
• [ACL 25] Text2World: Benchmarking Large Language Models for Symbolic World Model Generation, Webpage, Code
• [CoRL 25 Workshop] Articulated Object Manipulation using Online Axis Estimation with SAM2-Based Tracking, Webpage / Code
• [ECCV 24 Workshop (Best Paper)] RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins (early version), Webpage / Code

Co-Author

• [ICRA 26] LeHome: A Simulation Environment for Deformable Object Manipulation in Household Scenarios, TODO
• [ICLR 26] CubeBench: Diagnosing Interactive, Long-Horizon Spatial Reasoning Under Partial Observations, arXiv
• [ICLR 26] SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning, arXiv
• [ICLR 25 Workshop] HyCodePolicy: Hybrid Language Controllers for Multimodal Monitoring and Decision in Embodied Agents, arXiv
• [ICLR 26] AutoBio: A Simulation and Benchmark for Robotic Automation in Digital Biology Laboratory, Code
• [ICRA 26] AVR: Active Vision-Driven Robotic Precision Manipulation with Viewpoint and Focal Length Optimization
• [RSS 25] CordViP: Correspondence-based Visuomotor Policy for Dexterous Manipulation in Real-World, Webpage
• [CVPR 25] DexHandDiff: Interaction-aware Diffusion Planning for Adaptive Dexterous Manipulation, Webpage
• [ACL 25] HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model, arXiv
• [arXiv 24] ManiCM: Real-time 3D Diffusion Policy via Consistency Model for Robotic Manipulation: Webpage / Code
• [ICARM 24] Robust UAV Policy Learning for Urban Infrastructure Surface Screening, Paper

📷 Open-source Projects

• Embodied-AI-Guide 具身智能技术指南: Code
• Awesome-Embodied-AI-Job 具身智能招贤榜: Code

Collapse