# US 2022/0284641 A1 · Systems and methods of generating color palettes with a generative adversarial network > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2022/0284641 A1 - Document ID: US20220284641A1 - Status: Published application - Kind: A1 - Application number: 17/751,862 - Filed: 2022-05-24 - Published: 2022-09-08 - Earliest claimed: 2020-11-06 - Assignees: Salesforce, Inc. - Applicants: Salesforce, Inc. - Inventors: Michael Sollami; Amir Hossein Raffiee; Owen Winne Schoppe - Web page: https://patentgrep.com/patent/US20220284641A1 ## CPC - G06T 11/10: Texturing; Colouring; Generation of textures or colours (retouching, inpainting or scratch removal G06T5/77) - G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting - G06F 18/23: Clustering techniques - G06T 7/90: Determination of colour characteristics - G06V 10/751: {Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching} - G06T 2200/24: involving graphical user interfaces [GUIs] - G06T 2207/10024: Color image - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract Systems and method are provided for generating, at a server of a generative adversarial network (GAN) for color selection, a training set of color palettes. A color palette generator of the server generates a first set of color palettes based on the training set of color palettes. The first set of color palettes may be compared with a reference set of color palettes to predict a curated set of color palettes. Colors from the curated set of color palettes may be removed that are within a predetermined distance from one another in a color space. The GAN may be validated by performing cluster analysis to determine outlier latent dimensions to be changed for the color selection by the GAN. Proposed color palettes may be generated based on the GAN to be displayed on a display device. ## Claims ### Claim 1 A method comprising: generating, at a server of a generative adversarial network (GAN) for a color selection, a training set of color palettes; removing, at the server, colors from a curated set of color palettes that are within a predetermined distance from one another in a color space; validating, at the server, the GAN by performing a cluster analysis to determine outlier latent dimensions to be changed for the color selection by the GAN; and generating, at the server using the validated GAN, proposed color palettes to be displayed on a display device. ### Claim 2 (depends on 1) The method of claim 1, wherein the validating further comprises: inspecting, at the server, the determined outlier latent dimensions; and changing one or more of the dimensions for the color selection by the GAN based on the inspection. ### Claim 3 (depends on 2) The method of claim 2, wherein the inspecting the determined outlier latent dimensions further comprises: generating, at the server, cluster maps to determine the outlier dimensions. ### Claim 4 (depends on 1) The method of claim 1, wherein the generating the proposed color palettes further comprises: generating, at a color palette generator of the server, the proposed color palettes based on a color palette to be completed and random input color sequences. ### Claim 5 (depends on 1) The method of claim 1, wherein the generating the proposed color palettes further comprises: generating, at a color palette generator of the server, the proposed color palettes based on at least one received color characteristic and random input color sequences, wherein the at least one received color characteristic is selected from the group consisting of: hue, saturation, brightness, or color temperature. ### Claim 6 A system comprising: a server having a processor and a memory to store a generative adversarial network (GAN) for color selection to: generate, at a server of a generative adversarial network (GAN) for a color selection, a training set of color palettes; remove, at the server, colors from a curated set of color palettes that are within a predetermined distance from one another in a color space; validate, at the server, the GAN by performing a cluster analysis to determine outlier latent dimensions to be changed for the color selection by the GAN; and generate, at the server using the validated GAN, proposed color palettes to be displayed on a display device. ### Claim 7 (depends on 6) The system of claim 6, wherein the server validates the GAN by inspecting the determined outlier latent dimensions, and changes one or more of the dimensions for the color selection by the GAN based on the inspection. ### Claim 8 (depends on 7) The system of claim 7, wherein the server inspects the determined outlier latent dimensions by generating cluster maps to determine the outlier dimensions. ### Claim 9 (depends on 6) The system of claim 6, wherein a color palette generator of the server generates the proposed color palettes based on a color palette to be completed and random input color sequences. ### Claim 10 (depends on 6) The system of claim 6, wherein the color palette generator generates the proposed color palettes based on at least one received color characteristic and random input color sequences, and wherein the at least one received color characteristic is selected from the group consisting of: hue, saturation, brightness, or color temperature.