Google patents 3D view synthesis using transformer models
Google has filed a patent application for a method that creates new 3D views of an environment from a single input image. The system aligns different perspectives in a shared 3D space and uses a transformer model to generate a new viewpoint that was not originally captured.
In the filing’s own words
A method for view synthesis includes aligning a plurality of representations of a first perspective and a second perspective within a three-dimensional space common to the first perspective and the second perspective and based on a mapping between the first perspective and the second perspective. The method includes transforming the plurality of representations into a first set of descriptors representing the first perspective and a second set of descriptors representing the second perspective and configuring a model based on the first set of descriptors and the second set of descriptors. A third perspective of a second environment is generated based on an input image of the second environment.
What exactly they claim (claim 1)
1. A method comprising: aligning a plurality of representations of a first perspective in a first environment and a second perspective in the first environment within a three-dimensional space common to the first perspective and the second perspective and based on a mapping between the first perspective and the second perspective; transforming the plurality of representations into a first set of descriptors representing the first perspective and a second set of descriptors representing the second perspective; configuring a model based on the first set of descriptors and the second set of descriptors; and generating a third perspective of a second environment based on an input image of the second environment of the second environment and the model.
Facts
- Company
- Google (GOOGLE LLC)
- Official title
- TOKEN-DISENTANGLED TRANSFORMER ARCHITECTURES FOR VIEW SYNTHESIS
- Publication number
- US20260268589A1
- Application number
- 19559606
- Filed
- Mar 6, 2026
- Published
- Sep 10, 2026
- Inventors
- 5
- Claims
- 21
Classified under
G06T 15/205 Physics & computingG06T 3/18 Physics & computingG06T 7/344 Physics & computingG06T 7/75 Physics & computingMore from Google in AI & software
- Google patents video search that answers questions using AI descriptions Sep 15, 2026
- Google patents photo tool that removes people and their gear Sep 15, 2026
- Google patents AI that writes live meeting notes and summaries Sep 15, 2026
- Google patents AI tool that turns messy emergency notes into clear messages Sep 3, 2026
- Google patents AI that clicks through apps to finish tasks for you Sep 24, 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.