Stability AI, an AI developer, has created a technology based on generative AI that enables lightning-fast 3D image generation. Whereas previously it required powerful hardware and complex programming, the Stable Fast 3D model, consisting of two billion parameters, can now generate vivid, photorealistic images based on text or reference in just half a second, reports VentureBeat.
Image source: Stability.ai
The company says the technology marks a significant breakthrough in 3D modeling by cutting image processing time. “This is 1,200 times faster than our previous model, Stable Video 3D, released in March, which took up to 10 minutes to create just one 3D object,” Stability AI says.
Stable Fast 3D promises to be a powerful tool for a variety of industries, including design, architecture, retail, virtual reality and game development. It is based on TripoSR technology, developed in collaboration with 3D modeling company Trip AI.
Image source: Stability.ai
In the research articlepublished by Stability AI, details the methods used by the new model to quickly reconstruct high-quality 3D meshes from regular 2D images. The idea is to increase the speed of generating the final result without losing quality. The technology uses an advanced neural network transformer to create high-resolution volumetric images without significantly increasing computing power, which reduces artifacts and produces more detailed 3D models.
In addition, Stable Fast 3D uses an innovative approach to evaluating lighting and materials. The AI model can determine global values, such as roughness, fluidity, or metal, using a probability method that improves image quality. The technology also allows combining several elements required for a 3D image, including mesh, textures, and material properties, into a compact, ready-to-use 3D asset.
Interestingly, Stability AI continues to actively expand the boundaries of generative AI, moving from 2D to 4D. The company, which began its journey with generating images from text – Stable Diffusion, released Stable 3D in November 2023. In March of this year, it debuted Stable Video 3D technology with the ability to basic camera panning for viewing images and improved quality of 3D image generation. And just last week, it announced Stable Video 4D technology, which adds a time dimension to the generation of short 3D videos.
The Stable Fast 3D model is available through Stability AI’s Stable Assistant chatbot, the Stability AI API, and under license from the Hugging Face research community.
If you notice an error, select it with your mouse and press CTRL+ENTER.
