Introducing MuJoCo 3.0: a major new release of our fast, powerful and open source tool for robotics research. 🤖 📈 GPU & TPU acceleration through #JAX 🖼️ Better simulation of more diverse objects - like clothes, screws, gears and donuts 💡 Find out more: mujoco.org/3

Imitating a large scale of MoCap data for biomimetic robots is indeed promising, as shown in our work, especially considering its cost-effectiveness compared to hand-designed rewards. It's exciting that Tesla Optimus is moving in this direction.

Excited about Meta's consistent release of awesome datasets! I believe this new one will help humanoid robots learn a diverse set of skills.

Our paper's code on NCP is out! Explore and let's push the boundaries of what's possible together. Find it here: tencent-roboticsx.github.io/NCP/

Using a codebook as the latent space is quite effective for character control and animation, as it is compact and more interpretable.

I'm thrilled to announce that our paper has been published in Nature Machine Intelligence Nature Machine Intelligence today! To encourage research on the natural and agile movement of quadrupedal robots, we've open-sourced our code and data. Check out our project page: tencent-roboticsx.github.io/lifelike-agili…

I highly encourage you to try our MoCap data to train your quadrupedal robot. If you're tired of designing numerous rewards or want your robot to move more agilely and efficiently, our data is perfect for you! It includes walking, running, jumping, sitting, and playing.

Excited to announce that our research has made the cover of Nature Machine Intelligence ! 🎉 Honored to have our work recognized in this way. Big thanks to the team and everyone who supported us!