IP Library Granted Patent US 12,108,100
Granted Patent B2
US 12,108,100 · App. 18/220,459 · Granted Oct 1, 2024

Systems and methods for generating adapted content depictions

Inventors: Deviprasad Punja (Bangalore, IN); Madhusudhan Srinivasan (Karnataka, IN); Alan Waterman (Merced, CA)
Assignee: Rovi Guides, Inc.
H04N21/251G06N3/044G06N3/08H04N21/25883H04N21/25891
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 12,108,100
App. No.
18/220,459
Granted
Oct 1, 2024
Kind
B2
Abstract

A method for generating a content depiction of particular content that includes a machine learning system programmed to receive profile data representing preferences for content. The machine learning system identifies preferences for content features based upon the profile data, accesses content data representing the particular content and other content, and classifies features of the content data and content structure data within a content structure database system according to content categories. The machine learning system generates a content structure depiction of the particular content by combining content structure data from the content structure database system, wherein the combining is based upon correlating the identified preferences of the profile with the classified content categories. The machine learning system receives feedback data responsive to the content depiction and reprograms a configuration of the machine learning system for generating a content depiction based upon the feedback data.

Claims (47)

1. A method comprising:

receiving image data;

receiving object data;

retrieving a plurality of verified depictions of content;

accessing a neural network framework, wherein the neural network framework comprises a generator neural network and a discriminator neural network;

generating a proposed content depiction, by inputting into the generator neural network of the neural network framework: at least one of the image data, the object data, or the plurality of verified depictions of content;

determining whether the proposed content depiction is accepted by inputting into the discriminator neural network of the neural network framework: the proposed content depiction generated by the generator neural network, and the plurality of verified depictions of content;

based on determining that the proposed content depiction is not accepted:

modifying configurations of the generator neural network of the neural network framework and the discriminator neural network of the neural network framework; and

repeating the generating and the determining steps;

based on determining the proposed content depiction is accepted:

storing the proposed content depiction in a content platform in association with the content; and

generating for display the stored proposed content depiction on the content platform.

2. The method of claim 1 , further comprising:

causing distribution of the proposed content depiction across a computer network to at least one network device.

3. The method of claim 1 , wherein the image data comprises at least one of an image, an image attribute, image metadata, image data obtained from object data, or image data obtained from a depiction.

4. The method of claim 1 , wherein the object data comprises at least one of an object name, an object type, an object feature, an object state, an object action, an object absolute location, an object relative location, an object absolute motion, or an object relative motion.

5. The method of claim 1 , wherein each of the plurality of verified depictions of content comprises at least one of a poster or an image.

6. The method of claim 1 , wherein the generator neural network has an input layer having nodes representing the at least one of the image data, the object data, or the plurality of verified depictions of content, and a processing layer of nodes and connections between them, the nodes and connections programmed and configured to output the proposed content depiction to an output layer.

7. The method of claim 6 , wherein the discriminator neural network is programmed to compare the proposed content depiction with features of at least one of the plurality of verified depictions of content.

8. The method of claim 7 , wherein the discriminator neural network comprises an input layer of nodes representing the proposed content depiction and the plurality of verified depictions of content, and a processing layer of nodes and connections between them, the nodes and connections programmed and configured to output a determination of whether the proposed content depiction satisfies criteria of an acceptable content depiction.

9. The method of claim 8 , wherein the generator neural network and discriminator neural network are trained by feedback data, and wherein the generator neural network is trained by the discriminator neural network for the determination of whether the proposed content depiction satisfies the criteria of the acceptable content depiction.

10. The method of claim 9 , wherein the feedback data comprises content consumption tracked in response to distribution of the proposed content depiction.

11. A system comprising:

circuitry configured to:

receive image data;

receive object data;

retrieve a plurality of verified depictions of content;

access a neural network framework, wherein the neural network framework comprises a generator neural network and a discriminator neural network;

generate a proposed content depiction, by inputting into the generator neural network of the neural network framework: at least one of the image data, the object data, or the plurality of verified depictions of content;

determine whether the proposed content depiction is accepted by inputting into the discriminator neural network of the neural network framework: the proposed content depiction generated by the generator neural network, and the plurality of verified depictions of content;

based on determining that the proposed content depiction is not accepted:

modify configurations of the generator neural network of the neural network framework and the discriminator neural network of the neural network framework; and

repeat the generating and the determining steps;

based on determining the proposed content depiction is accepted:

store the proposed content depiction in a content platform in association with the content; and

generate for display the stored proposed content depiction on the content platform.

12. The system of claim 11 , wherein the circuitry is configured to:

cause distribution of the proposed content depiction across a computer network to at least one network device.

13. The system of claim 11 , wherein the image data comprises at least one of an image, an image attribute, image metadata, image data obtained from object data, or image data obtained from a depiction.

14. The system of claim 11 , wherein the object data comprises at least one of an object name, an object type, an object feature, an object state, an object action, an object absolute location, an object relative location, an object absolute motion, or an object relative motion.

15. The system of claim 11 , wherein each of the plurality of verified depictions of content comprises at least one of a poster or an image.

16. The system of claim 11 , wherein the generator neural network has an input layer having nodes representing the at least one of the image data, the object data, or the plurality of verified depictions of content, and a processing layer of nodes and connections between them, the nodes and connections programmed and configured to output the proposed content depiction to an output layer.

17. The system of claim 16 , wherein the discriminator neural network is programmed to compare the proposed content depiction with features of at least one of the plurality of verified depictions of content.

18. The system of claim 17 , wherein the discriminator neural network comprises an input layer of nodes representing the proposed content depiction and the plurality of verified depictions of content, and a processing layer of nodes and connections between them, the nodes and connections programmed and configured to output a determination of whether the proposed content depiction satisfies criteria of an acceptable content depiction.

19. The system of claim 18 , wherein the generator neural network and discriminator neural network are trained by feedback data, and wherein the generator neural network is trained by the discriminator neural network for the determination of whether the proposed content depiction satisfies the criteria of the acceptable content depiction.

20. The system of claim 19 , wherein the feedback data comprises content consumption tracked in response to distribution of the proposed content depiction.

Assignments (2)
CHANGE OF NAME Recorded Oct 3, 2024
From: ROVI GUIDES, INC.
To: ADEIA GUIDES INC.
Reel/Frame 069106/0238 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 11, 2023
From: PUNJA, DEVIPRASAD; SRINIVASAN, MADHUSUDHAN; WATERMAN, ALAN
To: ROVI GUIDES, INC.
Reel/Frame 064217/0147 →