Hangxin Liu

Assistant Professor @PKU1898 | Working on #robotics
Ph.D. in Computer Science from @UCLA
B.S. in Mechanical Engineering and Computer Science from @virginia_tech

[Hangxin Liu @HangxinLiu]

🤖🤖🤖

[Siyuan Huang @siyuanhuang95]

We are building next-gen humanoid foundation models to achieve General Humanoid Intelligence, in both academia and industry levels!🤖

If you are obsessed with whole-body control, sim2real, dexterous manipulation, and robotics foundation models, come join us in Beijing! We have a lineup of outstanding researchers and engineers.

Hiring:
🚀 Research Interns
💼 Full-time Researchers & Engineers

DMs are open. Let’s build the future together. 🚀

[Hangxin Liu @HangxinLiu]

Check out our new paper: Cross-Robot Behavior Adaptation through Intention Alignment

We enable skill generalization across different embodiments and tasks by encoding both intention descriptions and robot motion generators into a shared latent space.

[Science Robotics @SciRobotics]

A new Science #Robotics study demonstrates that robots can reproduce movements demonstrated by peers by aligning their abilities with the task’s objective, instead of just copying motor movements directly. https://t.co/m5xuGh7Y84

[Siyuan Huang @siyuanhuang95]

🎉🎉 UniFP won the best paper award at CoRL25, we truly appreciate that the award committee and organizers really care about the fundamental control problem that can be solved by learning-based approach!

Meanwhile, I felt fully engaged during the entire conference, from keynote to oral to poster, which is a truly pleasure experience after many years research. The organization and vibe of #CoRL2025 are wonderful!

Congratulations to all the collaborators, the code will be fully open-sourced very soon!

[Namiko Saito @namikosaito_]

📣Call for participation to our workshop in IROS2025 at Hangzhou, China

The Art of Robustness: Surviving Failures in Robotics🎨🤖
https://t.co/4Pe89vgbeX

Workshop date: Oct 20th AM
Submittion deadline: Sep 20th

[Hangxin Liu @HangxinLiu]

Our latest publication in Nature Machine Intelligence presents a dexterous hand embedded with 17 vision-based tactile sensors. Dexterousmanipulation and rich tactile sensing are key milestones on the path to truly capable #HumanoidRobot. 🤖🖐️
🔗 https://t.co/sVja9bQaR9

[Hangxin Liu @HangxinLiu]

5 km run, zero mistakes. 🏃‍♂️🤖 Warming up for the Beijing Robot Half Marathon. #RobotMarathon

[Hangxin Liu @HangxinLiu]

A joint effort between our lab and Leju Robotics. No video edits—just raw performance! A single 2-min shot isn’t perfect for validation, but it’s way better than a mashup of 5-sec clips. Show how the robot really operates! 🚀 #Robotics #AI

[The Humanoid Hub @TheHumanoidHub]

Kuafu humanoids by Chinese company Leju Robotics moving totes at FAW's car factory.

[Hangxin Liu @HangxinLiu]

Getting used to a humanoid co-worker? 🤖#Humanoid

[Xiaojian Ma @jeasinema]

📢 Attending #NeurIPS2024 ? Come by our workshop on open-world agents!

everything 👉 https://t.co/UpvUNGDIF4

Put your questions for the panel here: https://t.co/Yhq3O3Qo5w

Our speakers & panelists lining up: @mengjiao_yang @taoyds @xiao_ted @natashajaques @jiajunwu_cs @JohnCLangford @9LdROhjZE56jSh9

[Ruiqi Gao @RuiqiGao]

A common question nowadays: Which is better, diffusion or flow matching? 🤔

Our answer: They’re two sides of the same coin. We wrote a blog post to show how diffusion models and Gaussian flow matching are equivalent. That’s great: It means you can use them interchangeably.

[Bipasha Sen @bipashasen31]

🚀 Introducing Frank 🤖—a whole-body robot control system for day-to-day household chores.

Frank has been in works for the past year—a tightly coupled hardware and a remote teleoperation interface co-led by @michl_wang, @nst372, and @pulkitology.

Thanks to @skymanaditya1 and @shankaravaidy for the awesome help with recording the videos! 🎥

🧵[1/6]

[Brett Adcock @adcock_brett]

@engineered_arts Chinese startup Leju Robotics released an open-source humanoid development platform for academic and R&D use.

It includes an SDK for sensors and controls, simulation models, an LLM interface, and basic demos.

A great resource for robotics researchers!

[Siyuan Huang @siyuanhuang95]

📢📢This year our group has four papers to present at ECCV24. They tackled the key problems in 3D reconstruction, grounding, and interaction with models and datasets!

🤗🤗We have arrived at Milan, welcome to chat with us, and excited to meet old and new friends!

[Hangxin Liu @HangxinLiu]

Welcome to contribute your best ideas and submissions. And do stop by the workshop for a chat!

[Xiaojian Ma @jeasinema]

📣📣We're pleased to unveil the 1st Workshop on Open-World Agents (OWA-2024) along with #neurips2024 in stunning Vancouver, BC, Canada!🏔️🍁

🏆1 best paper award (2 runners-up), 4 orals
⏳submission due: 09/20/2024
🌐 website: https://t.co/UpvUNGDIF4

Incredible speakers & panelists lining up, see below👇
@mengjiao_yang @taoyds @xiao_ted @natashajaques @jiajunwu_cs @JohnCLangford

[Hangxin Liu @HangxinLiu]

🚨 Excited to share our latest oral paper at #IROS2024! 🚨
We’ve designed a modular GelSight sensor system that scales—and built various forms of gripper with it!

paper👉https://t.co/28m9dRBaDh

[Zilong Zheng @ZilongZheng]

📢📢Ready for #acl2024. Our team will present five papers covering topics of RL4NLP, long context, LLM agents, etc. We will also discuss recent insights on EmbodiedAI @ SpLU-RoboNLP. Code and datasets are all released. Welcome to chat over the posters, DMs are open! #aclmeeting

[Hangxin Liu @HangxinLiu]

Checkout the LLM + TAMP from https://t.co/emQdxDrN4j by @ShuWang_UCLA!!

[Shu Wang @ShuWang_UCLA]

🎉Thrilled to share that LLM3 has been accepted to #IROS2024!

LLM3: Large Language Model-based Task and Motion Planning with Motion Failure Reasoning #LLMAgent #Robotics

arxiv: https://t.co/wTBINoXPnK
code: https://t.co/54EzX1ELWz

[Siyuan Huang @siyuanhuang95]

🎉🎉🎉Super excited to announce that we launched a joint lab of Embodied AI and Humanoid Robot between UniTree and BIGAI!
I will direct the lab to foster the research in integrating the 3D scene understanding capability to Humanoid robot. Hope to see more generalist Humanoid robots in our home very soon😁

[Muzhi Han @muzhi_han]

🤖📜 Tired of hand-crafting predicates for your robot's task planning? Now you can teach predicates through natural language!

Introducing InterPreT, an LLM-powered framework that learns predicates and operators for robot task planning. #RSS2024

Website: https://t.co/uGWrZol3x6

[Zilong Zheng @ZilongZheng]

🔥🔥Checkout DiveR-CT, Diversity-enhanced Red Teaming with Relaxing Constraints

https://t.co/1XJwW4SfFC

Highlights:
🚩New optimization by constrained RL
🚩Marked superiority in diversity and ASR
🚩Allowing dynamic control of objective weights
🚩Enhancing resiliency in blue team