Color conversion using neural networks
US11023791B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Oct 30, 2019 |
| Grant date | Jun 1, 2021 |
| Priority date | — |
| Expiry date | Oct 30, 2039 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N3/047
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
An example system includes a processor and a non-transitory computer-readable medium having stored therein instructions that are executable to cause the system to perform various functions. The functions include obtaining a source profile associated with a print job and a destination profile associated with the print job. The functions also include, based on the source profile and the destination profile, generating a color conversion object that maps input colors of a source color space to output colors of a destination color space. In addition, the functions include training a neural network using the color conversion object so as to obtain weights associated with two or more hidden layers of nodes of the neural network. Further, the functions include receiving color data corresponding to pixels of the print job, and converting the color data from the source color space to the destination color space using the neural network.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.