<ul><li>Meta has unveiled a foundation model for controlling the behavior of virtual embodied humanoid agents, enabling them to learn human-like movements and perform complex tasks.</li><li>The model, called Meta Motivo, can adjust to changes in the environment such as gravity and wind without specific training for them.</li><li>Meta believes this research could lead to more lifelike NPCs, democratization of character animation, and new immersive experiences in the Metaverse.</li><li>In addition, Meta has launched Meta Video Seal, a framework for neural video watermarking to verify the origin of AI-generated content and mitigate misuse risks.</li></ul>

Meta Says Foundation Model Gives Virtual Embodied Agents Human-Like Movements

Discover more