Congrats on the amazing work! Tairan He @ICRA . The long-lasting myth of Sim2Real is getting quickly demystified every day by works like ASAP. 2025 might be the year that we finally crack it.

How do I see Figure and 1x demo from a data perspective? My first blog: scaling data collection for robotics foundation model. Teleop is not the end game. I bet on pre-training via 𝙬𝙤𝙧𝙡𝙙 𝙢𝙤𝙙𝙚𝙡 𝙖𝙣𝙙 𝙜𝙚𝙣𝙚𝙧𝙖𝙩𝙞𝙫𝙚 𝙨𝙞𝙢. Here is why: haoruxue.github.io/data-scaling-l…

Sim2Real RL for Vision-Based Dexterous Manipulation on Humanoids toruowo.github.io/recipe/ TLDR - we train a humanoid robot with two multifingered hands to perform a range of dexterous manipulation tasks robust generalization and high performance without human demonstration :D

This robot will rapidly become more capable over the next six months and it will seem to observers like we are accelerating, but in fact, most of the hard, risky work is done now. We've got software problems instead of hardware problems.

your project when you have all the robots in the world to spare:

Next-gen vision pre-trained models shouldn’t be short-sighted. Humans can easily perceive 10K x 10K resolution. But today’s top vision models—like SigLIP and DINOv2—are still pre-trained at merely hundreds by hundreds of pixels, bottlenecking their real-world usage. Today, we

Introducing COMPACT: COMPositional Atomic-to-complex Visual Capability Tuning, a data-efficient approach to improve multimodal models on complex visual tasks without scaling data volume. 📦 arxiv.org/abs/2504.21850 1/10

Nice perceptive locomotion on terrain. Very cool idea leveraging simple video data source!

Really nice agility & adaptivity

🤖 Can a humanoid robot hold extreme single-leg poses like Bruce Lee's Kick or the Swallow Balance? 🤸 💥 YES. Meet HuB: Learning Extreme Humanoid Balance 🔗 Project website: hub-robot.github.io

Cool dex manipulation pipeline in the wild 🔥

🏆 𝗕𝗲𝘀𝘁 𝗣𝗮𝗽𝗲𝗿 𝗙𝗶𝗻𝗮𝗹𝗶𝘀𝘁 for DIAL-MPC at #ICRA2025! Catch us at the award session: Wednesday 8:30am Room 302 We’ve also released a new demo: ball-spinning on finger tip. Enjoy 👉 github.com/LeCAR-Lab/dial…

Slides are here if you are interested: drive.google.com/file/d/1xta-H2…

Career Update: I’m interning at NVIDIA GEAR Lab supervised by Jim Fan and Yuke Zhu. Looking forward to frontier robot learning research with the magnificent team!

🤖Can a humanoid robot carry a full cup of beer without spilling while walking 🍺? Hold My Beer ! Introducing Hold My Beer🍺: Learning Gentle Humanoid Locomotion and End-Effector Stabilization Control Project: lecar-lab.github.io/SoFTA/ See more details below👇