SimpleAI releases HiFi-UMI: high-fidelity data engine enables direct real-robot deployment without teleoperation data
机器之心 · wechat · 2026-08-04
SimpleAI introduces HiFi-UMI, a high-fidelity data production engine for robot learning. It uses head-mounted stereo inertial SLAM, hardware synchronization, multi-view vision, and full-palm gloves to address issues like trajectory drift, inaccurate bimanual poses, time misalignment, and blind spots in traditional UMI data. Experiments show that policies trained solely on HiFi-UMI data can be deployed on real robots with success rates comparable to teleoperation data. Large-scale pretraining reduces action prediction error by 41% and shows power-law scaling. The team open-sourced the HiFi-UMI-2K dataset.
More from Embodied
- Humanoid Robot Production Could Reach 50M in 5 Years, Driven by Scaling Effects — willccbb · 2026-08-04
- Programming Robots With the Wrong Lab Music Shows — broodsugar · 2026-08-04
- Qianxun AI study: legacy data shows 'emergent transfer' after robot hardware upgrades, with a critical threshold — 机器之心 · 2026-08-04
- Stipple: Real-Time Incremental 3D Gaussian Splatting with Visual-Inertial Tracking — zhenjun_zhao · 2026-08-04
- Travis Kalanick's robotics startup Atoms raises $1.7B led by a16z — emmanuelvivier · 2026-08-04
- Building a Custom ESP32 Tamagotchi Easily with AI Agents — Scobleizer · 2026-08-04