Google DeepMind's Genie 2 Generates Interactive 3D Worlds from Single Images

Dec 4, 2024, 06:34 PM

Google DeepMind has introduced Genie 2, a new AI model that can generate interactive, playable 3D worlds from a single prompt image. This model, described as the company's most capable foundation world model, allows users to explore these environments for up to a minute using keyboard and mouse inputs. Genie 2 is designed to facilitate the training and evaluation of embodied AI agents by providing an endless variety of virtual environments. The model has the potential to revolutionize gaming, AI product testing, and the development of AI agents by offering a new tool for creating dynamic, interactive 3D spaces.