Pinscreen’s Avatar Creation SDK is the only solution for generating a complete personalized avatar from a single input photo. Simply by uploading a picture, we generate a fully rigged face, body model, and customized clothing within seconds - ready for animation and integration. Test it out by trying the Pinscreen app!
Our patented 3D digitization technology is the first and only solution that can generate a complete avatar head including a 3D hair model from a single input photo. Once digitized, you can personalize your look and change hairstyles regardless. Our hair model can adapt to any head shape automatically, and even hats can be simulated on them.
We have introduced a cutting-edge neural rendering technology called paGAN, which stands for photoreal avatar GAN. This method allows us to synthesize photorealistic 3D faces from a single 2D photo of a person. Our technology can be integrated on mobile device and is compatible with popular game engines such as Unreal and Unity.
In addition to the high-fidelity 3D head models generated from our AvatarSDK, we also provide fully rigged body models for both male and female subjects. The entire full body avatar can be loaded directly into your game engine or exported as an FBX format. We also provide body shape customization options with a highly diverse set of skin tones to handle any type of ethnicity.
Our image generated 3D avatars models come with a complete animation rig that is compatible with animation tools (e.g., Maya, Blender, etc.) and game engines (Unreal, Unity). For facial animation, our system generates 51 personalized blendshapes based on FACS, as well as eye gaze control. Our fully skinned body models come with 89 joints for male subjects and 86 joints for female ones.
• 3D Face Model including Blendshape rig
• Polystrip hair model including simulation layer
• Clothed body model including skeleton rig
• Texture and shading assets
• Basic animation package
Automatic Head and Hair Digitization from a Single Photo
Photoreal Facial Expression Synthesis
High performance: 10 seconds to generate an avatar once picture is uploaded
Integration with fully rigged body models (male and female)
Multiple levels of detail (low, medium, high resolution avatars)
Robust to challenging input photos (bad lighting conditions, complex facial expressions/poses)