🚀 Looking for a reliable Franka controller with seamless sim-to-real transfer?
Meet FrankaTwin: https://t.co/jbCOWbFp9g
🎮 Joint/Cartesian position & impedance + gripper
🎯 System ID: 3.6 mm EE / 28 mrad joint RMSE (sim vs. real)
What’s inside?
🎮 Joint/Cartesian position & impedance + gripper.
🧩 Easily set up in Isaac Sim for policy training and evaluation. Simple yet effective sysid.
📖 Clean codebase with a detailed document.
Huge thanks to @patrickhyin and Harry H. for their contributions!
🚀 Looking for a reliable Franka controller with seamless sim-to-real transfer?
Meet FrankaTwin: https://t.co/jbCOWbFp9g
🎮 Joint/Cartesian position & impedance + gripper
🎯 System ID: 3.6 mm EE / 28 mrad joint RMSE (sim vs. real)
It matches the task-space impedance control scheme used in sim-to-real research such as OmniReset, IndustReal, and our latest paper.
📄 IndustReal: https://t.co/nllciuFBtC
📄 OmniReset: https://t.co/HgfwcJJdOc
📄 Our paper (RebarSim): https://t.co/2KEskF0y1t
Can robots replace humans in dirty and hazardous construction work?
Introducing: Rebarbot 1.0🤖 https://t.co/tpPwBWsMRm
The first learning-based robotic system capable of autonomous rebar installation, marking an interesting step toward automated concrete component fabrication.
Introducing WoMAP—a new active perception framework merging world models with VLMs for open-vocabulary object localization in unknown environments. Grateful to have contributed a small part.
🔎Can robots search for objects like humans?
Humans explore unseen environments intelligently—using prior knowledge to actively seek information and guide search. But can robots do the same? 👀
🚀Introducing WoMAP (World Models for Active Perception): a novel framework for embodied open-vocabulary object localization that combines the reasoning power of VLMs 🧠with the physical grounding capabilities of world models 🌎.
🌐 https://t.co/e6ZUra86S1 🧵(1/N)
Think your RLHF-trained AI is aligned with your goals?
⚠️ We found that RLHF can induce significant misalignment when humans provide feedback by predicting future outcomes 🤔, creating incentives for LLM deception 😱
Introduce ✨RLHS (Hindsight Simulation)✨: By simulating future outcomes of the interaction before providing feedback, we drastically reduce misalignment and hallucinations 🦾