Industry
Training a robot to grasp an object means learning from many attempts, viewpoints and situations. One robot with three cameras at 20 fps pro…
Training a robot to grasp an object means learning from many attempts, viewpoints and situations. One robot with three cameras at 20 fps produces 216,000 images per hour. Now that robotics is scaling,
Training a robot to grasp an object means learning from many attempts, viewpoints and situations. One robot with three cameras at 20 fps produces 216,000 images per hour. Now that robotics is scaling, how can we efficiently train on all this data? https://x.com/i/article/2108539955777597440
Related
- Hugging Face Jobs sounds like an incredible addition to LeRobot🤗 https://huggingface.co/docs/lerobot/v0.6.0/hardware_guide#hugging-face-job…
- I think this means you can collect ~10k hours just from open source datasets, which means basically anyone should be able to build a decent …
- We built a bipedal robot for about $2,500. A real, mostly 3D-printed robot you can build, repair, simulate, train, and control. Today we’re …
Source: Clem Delangue (X) | 2026-10-09