Cewu Lu

FTP-1 not only excels on familiar tactile sensors but also achieves unprecedented success on unseen setups, redefining the potential for cross-sensor generalization in robotic manipulation.

SyVLA achieves unprecedented task success rates and generalization in real-world robotic applications by effectively decoupling intention from control.

A novel representation for articulated parts perception achieves 73% manipulation success without the need for extensive fine-tuning.

Forget generic retrieval signals – UniDoc-RL uses reinforcement learning to teach LVLMs how to actively perceive and reason about visual information, yielding a 17.7% performance boost.

Robots can now learn complex manipulation tasks directly from human demonstrations using only a pair of smart glasses, achieving zero-shot transfer without specialized hardware.

Robots can now perform contact-rich tasks with significantly improved success rates and reliability by explicitly reasoning about forces, outperforming prior methods by up to 48%.

Unlock human-like dexterity in robotic manipulation by combining RL-assisted teleoperation with a novel VLA architecture that leverages force and tactile feedback.

Imagine fixing your robot's mistakes *before* it even makes them: RoboPocket lets you train robots twice as efficiently using just your smartphone and AR.