Zixuan Chen
126
posts
@C___eric417
San Diego, CA
Joined August 2016
Pretty impressive!
Meet BFM-Zero: A Promptable Humanoid Behavioral Foundation Model w/ Unsupervised RL👉 lecar-lab.github.io/BFM-Zero/ 🧩ONE latent space for ALL tasks ⚡Zero-shot goal reaching, tracking, and reward optimization (any reward at test time), from ONE policy 🤖Natural recovery & transition
00:00
super cool! Looking forward to the paper!
Excited to share our latest progress on building Behavior Foundation Model for Humanoid Robots🎈 Forward roll, hip-pop dance, even cartwheel -- all the things you have never imagined the little G1 could do -- we have made it based on ONE model👌 Stay tuned for paper and code😉
00:00
Thanks Xiaolong for summarizing our work. Engineering is quite important in robotics. For motion tracking, even though DeepMimic was proposed by
@xbpeng4years ago, there are still tons of things to do to make it work on real robot.
This work is not about a new technique. GMT (General Motion Tracking) shows good engineering practices that you can actually train a single unified whole-body control policy for all agile motion, and it works in the real world, directly with sim2real without adaptation. This is
00:00
🚀Introducing GMT — a general motion tracking framework that enables high-fidelity motion tracking on humanoid robots by training a single policy from large, unstructured human motion datasets. 🤖A step toward general humanoid controllers. Project Website:
00:00
At the core of human skills is the coordination of whole-body. By doing whole-body teleoperation on robots, we can unleash the potential of coordinated movements and unlock the possibilities of collecting high-quality full-body robot data.
🤖Introducing TWIST: Teleoperated Whole-Body Imitation System. We develop a humanoid teleoperation system to enable coordinated, versatile, whole-body movements, using a single neural network. This is our first step toward general-purpose robots. 🌐humanoid-teleop.github.io
00:00