Create photorealistic 3D portraits from selfies
Generate 3D portraits from casual videos
Transform selfies into 3D avatars
Large Animatable Human Model
Create 3D portraits from selfies
Generate 3D portraits from casual videos
Generate 3D portraits from selfie videos
Turn casual videos into dynamic 3D portraits
3D Image Composer
Generate free-viewpoint portraits from videos
Scalable and Versatile 3D Generation from images
Generate customizable 3D portraits from videos
Paper for the LLM model known as ThorV2 by Floworks
Nerfies: Deformable Neural Radiance Fields is a cutting-edge technology designed to create photorealistic 3D portraits from 2D selfies. By leveraging Neural Radiance Fields (NeRF), it reconstructs 3D avatars with high fidelity, enabling applications in virtual meetings, gaming, and augmented reality. This innovative approach combines deformable modeling techniques to capture complex face shapes and expressions, making it highly versatile for 3D avatar creation.
• Photorealistic 3D Portraits: Generate incredibly lifelike 3D avatars from a single selfie. • Deformable Modeling: Captures dynamic face shapes and expressions, ensuring natural-looking avatars. • Efficient Reconstruction: Processes input images quickly to create detailed 3D models. • Customizable Outputs: Adjust lighting, pose, and expression to match your desired outcome.
What type of input does Nerfies require?
Nerfies works best with high-quality selfies taken under good lighting conditions. Avoid blurry or low-resolution images for optimal results.
How long does it take to create a 3D avatar?
Processing times vary depending on the complexity of the input, but typically takes a few seconds to a minute.
Can I use the 3D avatar in animations?
Yes, the generated 3D avatars can be used in animations by adjusting the pose and expression parameters during the customization step.