IP Library › Granted Patent US 12,731,308
Granted Patent B2
US 12,731,308 · App. 18/232,938 · Granted Sep 8, 2026

Generation and management of personalized images using machine learning technologies

Inventors: Ivan Babanin (London, GB); Aleksandr Belskikh (London, GB); Roman Golobokov (London, GB); Ivan Hladush (Lehi, UT); Volodymyr Kalinichenko (Lehi, UT); Hee Hun Kim (Los Angeles, CA); Diana Maksimova (London, GB); Alexandr Marinenko (Lehi, UT); Daniel Moreno Cuellar (New York, NY); Anton Pankratov (Edinburgh, GB); Hanna Rulevska (Almere, NL); Pavel Savchenkov (London, GB); Sergey Smetanin (London, GB); Jeremy Baker Voss (Los Angeles, CA)
Assignee: Snap Inc.
G06T11/23G06F3/0482G06T7/50G06T7/70
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,731,308
App. No.
18/232,938
Granted
Sep 8, 2026
Kind
B2
Abstract

Various examples described herein support or provide generation and management operations of personalized images using machine learning technologies, including receiving portrait images of an entity; using machine learning models to generate an identity that represents the entity; identifying a template that comprises text descriptions of a scene and conditions; and using a machine learning models to generate a personalized image based on the identity and the template.

Claims (62)

1 . A system comprising:

at least one processor;

at least one memory component storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:

receiving, from a device, a portrait image of an entity;

generating an identity that represents the entity by using the portrait image and a machine learning model, the machine learning model comprising a stable diffusion model;

causing display of a user interface on the device, the user interface comprising an invitation to generate a plurality of personalized images associated with a theme:

detecting an indication of a user selection of the invitation;

identifying a fixed number of templates associated with the theme for image generation based on a user account associated with the device, each of the fixed number of templates for image generation comprising a text description of a themed story scene and an image conditioning; and

using the machine learning model to generate a fixed number of personalized images based on the identity and the fixed number of templates.

2 . The system of claim 1 , wherein the personalized image comprises a personalized narrative image that illustrates the entity in a story scene.

3 . The system of claim 1 , wherein the condition comprises a visual resource that controls a generation of the personalized image, the visual resource comprising one of a sketch, a pose map, a depth map, a normal map, and a canny edge.

4 . The system of claim 1 , wherein the fixed number of templates are identified based on a random selection, and wherein the plurality of templates associated with the theme comprises a preselected template that is not subject to the random selection.

5 . The system of claim 1 , wherein the fixed number of templates is made available for the generating of the fixed number of personalized images without initiating a transaction process, wherein the user interface is a first user interface, wherein the invitation is a first invitation, and wherein the operations further comprise:

identifying a remaining set of templates from the plurality of templates associated with the theme; and

causing display of a second user interface on the device, the second user interface comprising a second invitation to unlock availability of the remaining set of templates.

6 . The system of claim 5 , wherein the operations further comprise:

detecting an indication of a user selection of the second invitation;

initiating the transaction process to unlock the availability of the remaining set of templates in exchange for an element; and

in response to detecting completion of the transaction process, using the machine learning model to generate a remaining set of personalized images based on the remaining set of templates and the identity that represents the entity.

7 . The system of claim 1 , wherein the operations further comprise:

causing display of the personalized image on a user interface of the device, the user interface comprising a user-selectable element that allows transmission of the personalized image to a further device associated with a further user account that is connected to the user account;

detecting an indication of a user selection of the user-selectable element; and

transmitting the personalized image as an ephemeral message to the further device.

8 . The system of claim 1 , wherein the user account is a first user account, and wherein the operations further comprise:

determining the template corresponds to a plurality of entities; and

identifying a plurality of user accounts that are connected to the first user account, each of the plurality of user accounts being associated with an identity that is generated based on one or more portrait images of a corresponding entity.

9 . The system of claim 8 , wherein the operations further comprise:

detecting an indication of a user selection of one or more entities from the plurality of entities; and

using the machine learning model to generate the personalized image based on the template and the one or more entities.

10 . A method comprising:

receiving, from a device, a portrait image of an entity;

generating an identity that represents the entity by using the portrait image and a machine learning model, the machine learning model comprising a stable diffusion model;

causing display of a user interface on the device, the user interface comprising an invitation to generate a plurality of personalized images associated with a theme;

detecting an indication of a user selection of the invitation;

identifying a fixed number of templates associated with the theme for image generation based on a user account associated with the device, each of the fixed number of templates for image generation comprising a text description of a themed story scene and an image conditioning; and

using the machine learning model to generate a fixed number of personalized images based on the identity and the fixed number of templates.

11 . The method of claim 10 , wherein the personalized image comprises a personalized narrative image that illustrates the entity in a story scene.

12 . The method of claim 10 , wherein the condition comprises a visual resource that controls a generation of the personalized image, the visual resource comprising one of a sketch, a pose map, a depth map, a normal map, and a canny edge.

13 . The method of claim 10 , wherein the fixed number of templates are identified based on a random selection, and wherein the plurality of templates associated with the theme comprises a preselected template that is not subject to the random selection.

14 . The method of claim 10 , wherein the fixed number of templates is made available for the generating of the fixed number of personalized images without initiating a transaction process, wherein the user interface is a first user interface, and wherein the invitation is a first invitation, further comprising:

identifying a remaining set of templates from the plurality of templates associated with the theme; and

causing display of a second user interface on the device, the second user interface comprising a second invitation to unlock availability of the remaining set of templates.

15 . The method of claim 14 , further comprising:

detecting an indication of a user selection of the second invitation;

initiating the transaction process to unlock the availability of the remaining set of templates in exchange for an element; and

in response to detecting completion of the transaction process, using the machine learning model to generate a remaining set of personalized images based on the remaining set of templates and the identity that represents the entity.

16 . The method of claim 10 , further comprising:

causing display of the personalized image on a user interface of the device, the user interface comprising a user-selectable element that allows transmission of the personalized image to a further device associated with a further user account that is connected to the user account;

detecting an indication of a user selection of the user-selectable element; and

transmitting the personalized image as an ephemeral message to the further device.

17 . The method of claim 10 , wherein the user account is a first user account, further comprising:

determining the template corresponds to a plurality of entities;

identifying a plurality of user accounts that are connected to the first user account, each of the plurality of user accounts being associated with an identity that is generated based on one or more portrait images of a corresponding entity;

detecting an indication of a user selection of one or more entities from the plurality of entities; and

using the machine learning model to generate the personalized image based on the template and the one or more entities.

18 . A non-transitory computer-readable storage medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising:

receiving, from a device, a portrait image of an entity;

generating an identity that represents the entity by using the portrait image and a machine learning model, the machine learning model comprising a stable diffusion model;

causing display of a user interface on the device, the user interface comprising an invitation to generate a plurality of personalized images associated with a theme;

detecting an indication of a user selection of the invitation;

identifying a fixed number of templates associated with the theme for image generation based on a user account associated with the device, each of the fixed number of templates for image generation comprising a text description of a themed story scene and an image conditioning; and

using the machine learning model to generate a fixed number of personalized images based on the identity and the fixed number of templates.

Assignments (3)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 13, 2025
From: BABANIN, IVAN; BELSKIKH, ALEKSANDR; GOLOBOKOV, ROMAN; MAKSIMOVA, DIANA; PANKRATOV, ANTON; SAVCHENKOV, PAVEL; SMETANIN, SERGEY
To: SNAP GROUP LIMITED
Reel/Frame 072551/0140 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 13, 2025
From: SNAP GROUP LIMITED
To: SNAP INC.
Reel/Frame 072551/0272 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 25, 2023
From: HLADUSH, IVAN; KALINICHENKO, VOLODYMYR; KIM, HEE HUN; MARINENKO, ALEXANDR; MORENO CUELLAR, DANIEL; RULEVSKA, HANNA; VOSS, JEREMY BAKER
To: SNAP INC.
Reel/Frame 065333/0971 →
Continuity (1)
Related Publication 20250054210A1 · Feb 13, 2025
References Cited (21)
US 11169675B1 · Anvaripour · 2021 [cited by examiner]
US 11288880B2 · Shaburov · 2022 [cited by examiner]
US 20070257920A1 · Neider · 2007 [cited by examiner]
US 20140041056A1 · Stoop · 2014 [cited by examiner]
US 20150193829A1 · Mukherjee · 2015 [cited by applicant]
US 20210385179A1 · Heikkinen · 2021 [cited by examiner]
US 20220319075A1 · Hu et al. · 2022 [cited by applicant]
US 20240265204A1 · Meeks · 2024 [cited by examiner]
WO WO2025038351A1 · 2025 [cited by applicant]
Jeong, Hyeonho, Gihyun Kwon, and Jong Chul Ye. “Zero-shot generation of coherent storybook from plain text story using diffusion models.” arXiv preprint arXiv:2302.03900 https://doi.org/10.48550/arXiv.2302.03900 (Year: … [cited by examiner]
Jin, Ze, and Zorina Song. “Generating coherent comic with rich story using ChatGPT and Stable Diffusion.” arXiv preprint arXiv: 2305.11067 https://doi.org/10.48550/arXiv.2305.11067 (Year: 2023). [cited by examiner]
Ruiz, Nataniel, et al. “DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation.” arXiv preprint arXiv:2208.12242v2 (Mar. 2023). (Year: 2023). [cited by examiner]
“Absolutely Crazy Ending to MTG Arena Sealed Open Day One”, [Online]. Retrieved from the Internet: <URL: https://youtu.be/2yxVAyRK7RA?t=15>, (Published May 20, 2023), 3 pgs. [cited by applicant]
“Alternative Text”, Office for Digital Accessibility (ODA), University of Minnesota, [Online]. Retrieved from the Internet: <URL: https://web.archive.org/web/20250114041208/https://accessibility.umn.edu/getting-started/… [cited by applicant]
“Contrast Checker”, WebAIM, web accessibility in mind, (Archived on Jan. 31, 2023), 2 pgs. [cited by applicant]
“International Application Serial No. PCT/US2024/041262, International Search Report mailed Nov. 15, 2024”, 4 pgs. [cited by applicant]
“International Application Serial No. PCT/US2024/041262, Written Opinion mailed Nov. 15, 2024”, 7 pgs. [cited by applicant]
“Make Apps More Accessible”, [Online]. Retrieved from the Internet: <https://web.archive.org/web/20220318041934/https://developer.android.com/guide/topics/ui/accessibility/apps>, (Archived on Mar. 18, 2022), 6 pgs. [cited by applicant]
“Supporting VoiceOver in Your App”, [Online]. Retrieved from the Internet: <https://developer.apple.com/documentation/uikit/supporting-voiceover-in-your-app>, (Retrieved on Mar. 18, 2025), 6 pgs. [cited by applicant]
“Text Legibility”, Material Design, [Online]. Retrieved from the Internet: <URL: https://web.archive.org/web/20230107134917/https://m2.material.io/design/color/text-legibility.html#legibility-standards>, (Archived on Ja… [cited by applicant]
Jing, Shi, et al., “InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning”, Arxiv. Org, Cornell University Library, 201 Olin Library Cornell University Ithaca, NY 14853, (Apr. 6, 2023). [cited by applicant]