{"slug": "encord-pairs-muscle-signals-with-video-to-train-robots-through-hand-occlusion", "title": "Encord pairs muscle signals with video to train robots through hand occlusion", "summary": "Encord released EMG-1 on October 20th, an open dataset that pairs forearm muscle signals with first-person video and optical motion capture to train robotic AI on hand motion that cameras miss, according to a PR Newswire announcement. Encord co-founder and co-CEO Eric Landau is extending the company's data business into biological signals, alongside co-founder Ulrik Stig Hansen and a new research division. The dataset's value will depend on whether outside researchers can reproduce the gains and apply them to real deployments.", "body_md": "# Encord pairs muscle signals with video to train robots through hand occlusion\n\n**Co-founders Eric Landau and Ulrik Stig Hansen are adding biological data to Encord's robotics work with an open dataset and a new research division.**\n\n        By [RuntimeWire Staff](https://runtimewire.com/author/runtimewire-staff)\n        · Published \n\nPrimary source: [PR Newswire](https://www.prnewswire.com/news-releases/encord-releases-emg-1-the-largest-open-dataset-linking-muscle-signals-to-robotic-ai-development-302914344.html)\n\n## Why it matters\n\nEncord is extending its data business into signals that cameras miss, positioning the company to serve robotics teams earlier in the training-data pipeline. The dataset's value will depend on whether outside researchers can reproduce the gains and apply them to real deployments.\n\n[Encord](https://encord.com/?ref=runtimewire) co-founder and co-CEO [Eric Landau](https://encord.com/author/eric-landau/?ref=runtimewire) is extending Encord's data business to a part of human movement that cameras miss: the electrical signals in the forearm that drive finger motion. On October 20th, [Encord released EMG-1](https://www.prnewswire.com/news-releases/encord-releases-emg-1-the-largest-open-dataset-linking-muscle-signals-to-robotic-ai-development-302914344.html?ref=runtimewire), an open dataset pairing those muscle signals with first-person video and optical motion capture…\n\nWhen you purchase through links in our articles, [we may earn a small commission](https://runtimewire.com/affiliate-monetization-standards). This doesn’t affect our editorial independence.", "url": "https://wpnews.pro/news/encord-pairs-muscle-signals-with-video-to-train-robots-through-hand-occlusion", "canonical_source": "https://runtimewire.com/article/encord-emg-1-muscle-signals-robot-training", "published_at": "2026-10-20 18:14:17+00:00", "updated_at": "2026-10-20 21:17:19.960423+00:00", "lang": "en", "topics": ["robotics", "artificial-intelligence", "machine-learning", "ai-research"], "entities": ["Encord", "Eric Landau", "Ulrik Stig Hansen", "EMG-1", "PR Newswire"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/encord-pairs-muscle-signals-with-video-to-train-robots-through-hand-occlusion", "markdown": "https://wpnews.pro/news/encord-pairs-muscle-signals-with-video-to-train-robots-through-hand-occlusion.md", "text": "https://wpnews.pro/news/encord-pairs-muscle-signals-with-video-to-train-robots-through-hand-occlusion.txt", "jsonld": "https://wpnews.pro/news/encord-pairs-muscle-signals-with-video-to-train-robots-through-hand-occlusion.jsonld"}}