I tried using the new AI 'Stable Fast 3D' that can create 3D models in 0.5 seconds.
Stability AI, known as the developer of Stable Diffusion, has released an AI model called ' Stable Fast 3D ' that can quickly generate 3D models from images. A free demo app was also provided, so I tried using it.
Introducing Stable Fast 3D: Fast 3D asset generation from a single image — Stability AI Japan
Stable Fast 3D is an AI model developed based on the technology of TripoSR , a 3D model generation AI announced in March 2024, and can generate a 3D model in 0.5 seconds with a GPU equipped with 7GB of VRAM.
There is already a free demo app available on the web that allows you to use Stable Fast 3D, so I will try to create a 3D model of the following image.
First, click the link below to access the demo web app.
Stable Fast 3D - a Hugging Face Space by stabilityai
https://huggingface.co/spaces/stabilityai/stable-fast-3d
Next, drag and drop the image you want to turn into a 3D model into the area marked 'Drop image here.'
Next, click 'Remove Background' to automatically remove the background.
Once you have deleted the background, click 'Run'.
The 3D model was generated in an instant.
The 3D model can be moved around in the browser, and when you look at the side view, you can see that it is able to accurately guess and generate parts that are not included in the image.
The back looks like this.
I will also try to make a 3D model of the following image.
The result looks like this.
If you change the angle, you can see that the 3D model is able to maintain the white, green, and blue stripes.
The model data of Stable Fast 3D is available at the following link. The license is 'Stability AI Community License', and individuals or organizations with annual revenue of less than $ 1 million (approximately 150 million yen) can use it commercially.
stabilityai/stable-fast-3d · Hugging Face
https://huggingface.co/stabilityai/stable-fast-3d
Related Posts:
in Review, Software, Web Application, Posted by log1o_hf