Nvidia patents method to animate 3D head avatars from a single photo
Nvidia describes a system that creates a 3D model of a person's head from a single 2D photo. It combines the person's unique facial features with a new expression from a different image to generate a realistic, animated portrait without needing extra training.
In the filing’s own words
Systems and methods are disclosed that animate a source portrait image with motion (i.e., pose and expression) from a target image. In contrast to conventional systems, given an unseen single-view portrait image, an implicit three-dimensional (3D) head avatar is constructed that not only captures photo-realistic details within and beyond the face region, but also is readily available for animation without requiring further optimization during inference. In an embodiment, three processing branches of a system produce three tri-planes representing coarse 3D geometry for the head avatar, detailed appearance of a source image, as well as the expression of a target image. By applying volumetric rendering to a combination of the three tri-planes, an image of the desired identity, expression and pose is generated.
What exactly they claim (claim 1)
1. A computer-implemented method, comprising: receiving a source image depicting a first subject with a first expression; extracting a three-dimensional (3D) shape from the source image to produce an identity surface associated with the first subject, wherein the first expression is replaced with a second expression; processing the source image to compute an appearance representation associated with the first subject; integrating the appearance representation and the identity surface to produce an identity representation that, when rendered generates an output image of the first subject with the second expression; processing a target image depicting a second subject with a target expression to produce an expression representation; integrating the expression representation with the identity representation to produce a modified identity representation; and rendering the modified identity representation to produce a second output image depicting the first subject with the target expression.
Facts
- Company
- Nvidia (NVIDIA Corporation)
- Official title
- Neural head avatar construction from an image
- Patent number
- US12743841B2
- Application number
- 18653723
- Filed
- May 2, 2024
- Granted
- Sep 22, 2026
- Inventors
- 6
- Claims
- 23
Classified under
G06T 15/08 Physics & computingG06T 13/40 Physics & computingG06V 10/82 Physics & computingG06V 40/174 Physics & computingG06T 2210/56 Physics & computingG06V 10/774 Physics & computingMore from Nvidia in AI & software
- Nvidia patents AI coach that analyzes your gameplay to suggest strategies Oct 13, 2026
- Nvidia patents AI system that writes and voices character conversations Sep 10, 2026
- Nvidia patents system that uses AI to find bugs in game videos Sep 17, 2026
- Nvidia patents AI method to make digital characters move realistically Sep 8, 2026
- Nvidia patents AI system that builds image generators from text prompts Sep 15, 2026
Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.