Gradio demo of CharacterGen (SIGGRAPH 2024)
Convert images to 3D models
Reconstruct 3D Gaussians from unposes images.
Create a 3D model from an image using depth mapping
Generate 3D models and videos from images
Create recursive 3D polygons and mathematical surfaces
Generate a dynamic 3D scene with rotating lights and knots
Create a dynamic and colorful 3D scene with random objects and lights
Generate 3D recursive polygons and math functions
Create a recursive 3D DNA strand visualization
Scalable and Versatile 3D Generation from images
Generate textured 3D models from text descriptions
Create 3D reconstructions from images using MASt3R and 3DGS
CharacterGen is a state-of-the-art tool demonstrated at SIGGRAPH 2024, designed to generate 3D characters from single images. It leverages advanced AI technology to create detailed and realistic 3D models from 2D input, making it a powerful resource for 3D modeling and character design.
What file formats does CharacterGen support for input and output?
CharacterGen accepts standard image formats (e.g., PNG, JPEG) for input. The output 3D models are provided in OBJ or GLB formats, ready for use in most 3D modeling software.
Can I customize the generated character's appearance?
Yes, you can adjust settings such as texture, clothing, and pose before generating the final model to achieve your desired result.
Is CharacterGen suitable for non-experts?
Absolutely! No prior 3D modeling experience is required. The tool's user-friendly interface makes it accessible to beginners and professionals alike.