Nature Will has launched its first foundational model for the physical world, Natura-1. Trained on 1 million hours of first-person video footage of humans interacting with diverse scenarios and objects, Natura-1 can handle unseen objects and tasks. It supports fine-grained instructions specifying the final position and orientation of objects, maintains control when grip methods change mid-action, performs distinct tasks with both hands, and tracks moved or opened objects during long tasks to determine the next step. This follows earlier news of Nature Will releasing the 50B-parameter MoE model Natura-1; this update provides additional details on its capabilities in the physical world.
Nature Launches Natura-1, the First Foundation Model for Physical World Tasks
TechFlowShare
Summary
Real-world assets (RWA) news breaks as Nature Will launches Natura-1, its first foundation model for physical world tasks. Trained on 1 million hours of first-person video data, the model handles unseen objects, executes fine-grained positioning, adapts during tasks, and tracks moved items. Inflation data volatility has not slowed development, with the model ready for real-world deployment.
Source:Show original
Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information.
Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.