Meta 3D Gen create 3D images from texts

Meta is expanding its capabilities in the field of artificial intelligence with the launch of a new 3D model generation platform called Meta 3D Gen.

This advanced technology promises to transform the way 3D assets are created, offering the generation of high-quality three-dimensional models from simple text commands. Check it out!

Read too:

Opera One R2 brings split-screen mode and new AI features
This is the first commercial created using Sora, OpenAI’s artificial intelligence

Technological Advancements of Meta 3D Gen

Meta 3D Gen, developed by Meta’s research department, uses a two-step approach to create detailed, textured 3D models. The first step generates 3D geometry from a text command, while the second step applies high-resolution textures to the model. According to Meta, this methodology results in superior quality compared to existing solutions, and is 3 to 10 times faster.

“This system can generate 3D assets with high-resolution textures and end-to-end material maps, with results that are superior in quality to previous state-of-the-art solutions — 3 to 10 times faster than previous work,” explains Meta AI on Threads.

Meta 3D Gen can create complete 3D models, with underlying mesh structures that support physically-based rendering (PBR). This means that the generated models are usable in real-world modeling and rendering applications, providing a powerful tool for designers, animators, and game developers.

The platform combines two of Meta’s core generative models, AssetGen and TextureGen, leveraging the strengths of each. Based on feedback from professional 3D artists, Meta claims that its new technology is preferred over other text-to-3D models “more often than not,” while offering significantly faster processing speeds.

Ideal use for entertainment and architecture

One of the great advantages of Meta 3D Gen is the ability to separate mesh models and texture maps, allowing for greater control over the final result. This enables iterative refinement, common in text-to-image generators, where users can adjust the input to the texture style without having to modify the assumed model.

This flexibility is essential for professionals seeking precision and quality in their projects, allowing fine adjustments that result in more realistic and detailed models.

Meta’s new 3D model generation technology is not only interesting and even a first for the field, it also represents a significant advance in the field of generative artificial intelligence. With the ability to create high-quality 3D models from simple text commands, Meta 3D Gen has the potential to revolutionize industries ranging from entertainment and gaming to architecture and design.

Meta continues to invest heavily in AI research and development, and the launch of Meta 3D Gen is proof of that.

For more details on the technology and its comparative results, Meta has made available a complete white paper that explores the nuances of Meta 3D Gen and its benefits over other solutions on the market.

Fonte: PetaPixel