RoboPapers: Dexterity & Manipulation
How robots learn to grasp, handle, and interact with the physical world.
The catalog of RoboPaper picks on the systems, and hard-won lessons behind robotic manipulation. From sensing and grasp planning to learned policies, tactile feedback, and generalization in the wild.
Tap to Preview

Human Universal Grasping

  • [Paper] - Human Universal Grasping

  • [Interview] - Ep#92: Human Universal Grasping

  • [Editorial] - Essential synthesis by GPT & RoboPapers hosts

  • [Code] - Human Universal Grasping

  • [Weights] - Human Universal Grasping (HUG)

  • [Datasets] - aria2mano

  • [Debate] - The future of humanoid grasping

    • DexUMI (Universal Manipulation Interface)

  • [Paper] - DexUMI: Using Human Hand as the Universal Manipulation Interface for Dexterous Manipulation

  • [Poster] - Dex Workshop

  • [Interview] - Ep#25: DexUMI: Using Human Hand as the Universal Manipulation Interface for Dexterous Manipulation

  • [Editorial Analysis] - Overview of DexUMI by GPT and RoboPaper Hosts

  • [Code] - DexUMI: Using Human Hand as the Universal Manipulation Interface for Dexterous Manipulation

  • [Deployment Guide] - DexUMI

  • [Hardware Guide] - DexUMI Assembly

    • Sim-to-Real Reinforcement Learning

  • [Paper] - Sim-to-Real Reinforcement Learning for Vision-Based Dexterous Manipulation on Humanoids

  • [Interview] - Ep#3 Sim-to-Real RL for Vision-Based Dexterous Manipulation on Humanoids

  • [Code] - Toru Lin sim-to-real RL

    • Cross-Embodiment Manipulation

  • Cross-Embodiment Robot Manipulation via a Unified Hand Action Space

  • Ep#97: Cross-Embodiment Robot Manipulation via a Unified Hand Action Space

    • Tutor Intelligence

  • [Interview] - Ep#85: Tutor Intelligence

  • [Blog] - Building a 100-Robot Data Factory Toward Factory-Ready AI

    • SAM2Act

  • Ep#1 SAM2Act

    • Performant Robotic Manipulation

  • Ep#58: RL-100: Performant Robotic Manipulation with Real-World Reinforcement Learning