Meta, the social media giant founded by Mark Zuckerberg, has unveiled its latest artificial intelligence (AI) model, 3D Gen, which can create textured 3D models from text cues. This new technology is a game-changer in the world of digital content creation as it allows for the rapid creation of high-quality shapes and textures in under a minute.
The Meta research team has introduced a new system called 3D Gen that utilizes two key components developed by the team: Meta 3D AssetGen and Meta 3D TextureGen. These tools allow for Text to 3D and Text to Texture Generation, respectively. By combining these components, 3D Gen can represent 3D objects in different spaces with a success rate of over 68 percent compared to single-stage models.
What sets 3D Gen apart from other industry standards is its ability to generate or retexture 3D objects at an unprecedented speed while maintaining exceptional fidelity and visual quality. Its support for physically based rendering (PBR) and generative retexturing of 3D shapes further enhances its realism.
This new technology offers significant benefits for AI-driven content creation in the digital space, providing a faster and more accurate approach to creating and retexturing 3D objects using text prompts. With its technical capabilities and superior performance compared to industry standards, Meta’s 3D Gen is set to revolutionize the world of digital content creation.
