Rendering a dynamic endemic banner on streaming platforms using content recommendation systems and advanced banner personalization

Patent metadata

Field	Value
Publication number	US-12598353-B2
Application number	US-202418901391-A
Country	US
Kind code	B2
Filing date	Sep 30, 2024
Priority date	Aug 17, 2022
Publication date	Apr 7, 2026
Grant date	Apr 7, 2026

Abstract

Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for utilizing a content recommendation system powering a streaming media publisher channel, in conjunction with an object recognition model, to enhance dynamic generation of a banner being shown to a user via an awareness or performance campaign. This method allows the platform to present the most relevant ML personalized in-channel content to the publisher platform users in endemic banners that run on the platform which then correspondingly helps drive user reach. An example embodiment operates by implementing personalized content banners that may act as a hook for channel users opening their streaming device, both active and lapsed, to enter back into the channel.

First claim

What is claimed is:

1 . A computer-implemented method for creating dynamic banners, the computer-implemented method comprising:
generating, by at least one computer processor and based on first metadata of a user profile, a first call to a content recommendation system powering a streaming media publisher channel for a first recommended content asset for a first target banner, wherein the first metadata includes a user's preferences for content assets; initiating, based on a trained machine learning model, an object recognition algorithm to identify a plurality of objects located within imagery of the first recommended content asset, wherein each object of the plurality of objects includes identifying second metadata; comparing the first metadata and the second metadata; selecting, based on metadata common to the first metadata and the second metadata, at least one object from the plurality of objects; extracting the at least one object from the imagery of the first recommended content asset; stitching the at least one object into the first target banner to form a first composite banner; generating, based on the second metadata, a second call to the content recommendation system for a second recommended content asset; stitching imagery of the second recommended content asset into a second target banner to form a second composite banner; and rendering the first and second composite banners on a display of a media device.

2 . The computer-implemented method of claim 1 , wherein the first and second composite banners each comprise an endemic banner.

3 . The computer-implemented method of claim 1 , wherein the media device comprises an Over-the-Top (OTT) device.

4 . The computer-implemented method of claim 1 , further comprising the content recommendation system instantiating the trained machine learning model, based on the first metadata of the user profile, to generate the first recommended content asset.

5 . The computer-implemented method of claim 1 , wherein the trained machine learning model comprises an image recognition model.

6 . The computer-implemented method of claim 5 , further comprising the image recognition model being trained to recognize any of: faces, themes, genres, scenes, media series, or text within the imagery of the first recommended content asset or the imagery of the second recommended content asset.

7 . The computer-implemented method of claim 1 , further comprising dynamically modifying one or more visual components of the first composite banner.

8 . The computer-implemented method of claim 7 , wherein the dynamically modifying one or more visual components comprises generating one or more of: artwork, movement, animation, cinemagraphs, resizing, scaling, cropping, image framing, color changes, font changes, or composite filling.

9 . The computer-implemented method of claim 1 , wherein the selecting comprises determining a closest match by the first metadata being identical to the second metadata, or at least a portion of the first metadata being identical to a portion of the second metadata.

10 . A system, comprising:
one or more memories; and at least one processor each coupled to at least one of the memories and configured to perform operations comprising: generating, based on first metadata of a user profile, a first call to a content recommendation system powering a streaming media publisher channel for a first recommended content asset for a first target banner, wherein the first metadata includes a user's preferences for content assets; initiating, based on a trained machine learning model, an object recognition algorithm to identify a plurality of objects located within imagery of the first recommended content asset, wherein each object of the plurality of objects includes identifying second metadata; comparing the first metadata and the second metadata; selecting, based on metadata common to the first metadata and the second metadata, at least one object from the plurality of objects; extracting the at least one object from the imagery of the first recommended content asset; stitching the at least one object into the first target banner to form a first composite banner; generating, based on the second metadata, a second call to the content recommendation system for a second recommended content asset; stitching imagery of the second recommended content asset into a second target banner to form a second composite banner; and rendering the first and second composite banners on a display of a media device.

11 . The system of claim 10 , where the first and second composite banners each comprise an endemic banner.

12 . The system of claim 10 , where the system comprises a streaming media device platform for an Over-the-Top (OTT) device.

13 . The system of claim 10 , the operations further comprising instantiating the trained machine learning model, based on the first metadata of the user profile, to generate the first recommended content asset.

14 . The system of claim 10 , wherein the trained machine learning model comprises an image recognition model.

15 . The system of claim 14 , the operations further comprising training the image recognition model to recognize any of: faces of actors, characters, or text within the imagery of the first recommended content asset or the imagery of the second recommended content asset.

16 . The system of claim 14 , the operations further comprising training the image recognition model to recognize any of: themes, genres, scenes, or a media series within the imagery of the first recommended content asset or the imagery of the second recommended content asset.

17 . The system of claim 10 , the operations further comprising dynamically modifying one or more visual components of the first composite banner by generating one or more of: artwork, movement, animation, cinemagraphs, resizing, scaling, cropping, image framing, color changes, font changes, or composite filling.

18 . A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
generating, based on first metadata of a user profile, a first call to a content recommendation system powering a streaming media publisher channel for a first recommended content asset for a first target banner, wherein the first metadata includes a user's preferences for content assets; initiating, based on a trained machine learning model, an object recognition algorithm to identify a plurality of objects located within imagery of the first recommended content asset, wherein each object of the plurality of objects includes identifying second metadata; comparing the first metadata and the second metadata; selecting, based on metadata common to the first metadata and the second metadata, at least one object from the plurality of objects; extracting the at least one object from the imagery of the first recommended content asset; stitching the at least one object into the first target banner to form a first composite banner; generating, based on the second metadata, a second call to the content recommendation system for a second recommended content asset; stitching imagery of the second recommended content asset into a second target banner to form a second composite banner; and rendering the first and second composite banners on a display of a media device.

19 . The non-transitory computer-readable medium of claim 18 , the operations furt

Assignees

• Roku Inc

Inventors

• Sanghavi Mehul
• Mahto Rohit
• Lee Kelly
• Taneja Madhulika

Classifications

• H04N21/4316

  for displaying supplemental content in a region of the screen, e.g. an advertisement in a separate window · CPC title

• H04N21/4667

  Processing of monitored end-user data, e.g. trend analysis based on the log file of viewer selections · CPC title

• H04N21/2668

  Creating a channel for a dedicated end-user group, e.g. insertion of targeted commercials based on end-user profiles {(information retrieval from the Internet by querying with filtering and personalisation G06F16/9535; arrangements for replacing or switching information during the broadcast H04H20/10; push services over packet-switching network H04L12/1859; adaptation of message content in packet-switching networks H04L51/063)} · CPC title

• H04N21/23418

  involving operations for analysing video streams, e.g. detecting features or characteristics (television picture signal circuitry for scene change detection H04N5/147; filtering for image enhancement G06T5/00; methods or arrangements for recognising scenes G06V20/00; arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59) · CPC title

• H04N21/812

  involving advertisement data (advertising per se G06Q30/02) · CPC title