# US 2023/0156284 A1 · Systems and methods for deep recommendations using signature analysis

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0156284 A1
- Document ID: US20230156284A1
- Status: Published application
- Kind: A1
- Application number: 17/969,950
- Filed: 2022-10-20
- Published: 2023-05-18
- Earliest claimed: 2019-11-27
- Assignees: Rovi Guides, Inc.
- Applicants: Rovi Guides, Inc.
- Inventors: Juan Gerardo Menendez
- Web page: https://patentgrep.com/patent/US20230156284A1

## CPC

- H04N 21/4668: {for recommending content, e.g. movies}
- G06F 18/2413: based on distances to training or reference patterns
- G06N 3/0499: Feedforward networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 20/00: Machine learning
- G06T 7/40: Analysis of texture (depth or shape recovery from texture G06T7/529)
- G06V 10/462: {Salient features, e.g. scale invariant feature transforms [SIFT]}
- G06V 10/48: by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/82: using neural networks
- G06V 20/46: {Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames}
- H04N 21/44008: {involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream (arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59)}
- H04N 21/4663: {involving probabilistic networks, e.g. Bayesian networks}
- H04N 21/4666: {using neural networks, e.g. processing the feedback provided by the user}
- H04N 21/8153: {comprising still images, e.g. texture, background image}
- G06T 2207/10016: Video; Image sequence
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Systems and methods are described herein for providing content item recommendations based on a video. Using feature vectors corresponding to at least one frame of a video (e.g., generated based on texture and shape intensity of a frame), a recommendation system improves content recommendation using analytic and quantitative characteristics derived from a frame of a content item rather than merely manually labeled bibliographic data (e.g., a genre or producer). The recommendation system may generate a feature vector based on a texture, a shape intensity (e.g., generated from a Generalized Hough Transform), and temporal data corresponding to at least one frame of a video. The feature vector is analyzed using a machine learning model (e.g., a neural network) to produce a machine learning model output. The recommendation system causes a recommended content item to be provided based on the machine learning model output.

## Claims

### Claim at position 1

-35. (canceled)

### Claim 36

A computer-implemented method, comprising:
determining a user interaction of a user profile with a first content item;
determining, based on the user interaction, the first content item is of interest to the user profile;
identifying a first scene of the first content item;
determining one or more scenes of a second content item that are similar to the first scene; and
generating for display to the user profile a recommendation of the second content item, based on determining that the one or more scenes of the second content item are similar to the first scene, wherein generating for display to the user profile the recommendation of the second content item comprises generating for display the one or more scenes of the second content item.

### Claim 37 (depends on 36)

The method of claim 36, wherein generating for display to the user profile the recommendation of the second content item further comprises generating for display an identifier of the second content item.

### Claim 38 (depends on 36)

The method of claim 36, further comprising:
determining one or more scenes of a third content item that are similar to the first scene; and
generating for simultaneous display to the user profile the recommendation of the second content item and a recommendation of the third content item, wherein generating for display to the user profile the recommendation of the third content item comprises generating for display the one or more scenes of the third content item.

### Claim 39 (depends on 36)

The method of claim 36, further comprising:
determining a texture associated with the first scene;
generating a shape intensity associated with the first scene;
generating a feature vector based on the texture, the shape intensity, and temporal data associated with the first scene;
analyzing the feature vector using a machine learning model to produce a machine learning model output; and
generating for display to the user profile the recommendation of the second content item based on the machine learning model output.

### Claim 40 (depends on 36)

The method of claim 36, wherein determining, based on the user interaction, the first content item is of interest to the user profile comprises:
determining that the first scene of the first content item is of interest to the user profile, based on input received from the user profile.

### Claim 41 (depends on 36)

The method of claim 36, wherein determining, based on the user interaction, the first content item is of interest to the user profile is based on a duration of the user interaction with the first content item and/or a duration of a user interaction with the first scene.

### Claim 42 (depends on 36)

The method of claim 36, further comprising:
generating for display an icon selectable to cause one or more other scenes similar to the first scene of the first content item, and/or similar to the one or more scenes of the second content item, to be generated for display.

### Claim 43 (depends on 36)

The method of claim 36, further comprising:
generating for simultaneous display the first scene of the first content item and the recommendation of the second content item, wherein the first scene is displayed during playing of a video corresponding to the first content item to the user profile.

### Claim 44 (depends on 36)

The method of claim 36, further comprising:
identifying a second scene of the first content item;
determining, based on the second scene, a third content item having one or more scenes similar to the second scene; and
generating for display to the user profile a recommendation of the third content item, based on determining that the one or more scenes of the third content item are similar to the second scene.

### Claim 45 (depends on 44)

The method of claim 44, wherein:
identifying the second scene of the first content item comprises determining that the second scene of the first content item is a current scene being displayed during playing of a video corresponding to the first content item to the user profile, the second scene being later in a duration of the first content item than the first scene;
the method further comprising:
generating for simultaneous display the second scene of the first content item and the recommendation of the third content item, and ceasing generating for display the recommendation of the second content item.

### Claim 46

A computer-implemented system, comprising:
memory;
control circuitry configured to:
determine a user interaction of a user profile with a first content item;
determine, based on the user interaction, the first content item is of interest to the user profile;
identify a first scene of the first content item;
determine one or more scenes of a second content item that are similar to the first scene; and
generate for display to the user profile a recommendation of the second content item, based on determining that the one or more scenes of the second content item are similar to the first scene, wherein the control circuitry generates for display to the user profile the recommendation of the second content item by generating for display the one or more scenes of the second content item.

### Claim 47 (depends on 46)

The system of claim 46, wherein the control circuitry is further configured to generate for display to the user profile the recommendation of the second content item by generating for display an identifier of the second content item.

### Claim 48 (depends on 46)

The system of claim 46, wherein the control circuitry is further configured to:
determine one or more scenes of a third content item that are similar to the first scene; and
generate for simultaneous display to the user profile the recommendation of the second content item and a recommendation of the third content item, wherein the control circuitry generates for display to the user profile the recommendation of the third content item by generating for display the one or more scenes of the third content item.

### Claim 49 (depends on 46)

The system of claim 46, wherein the control circuitry is further configured to:
determine a texture associated with the first scene;
generate a shape intensity associated with the first scene;
generate a feature vector based on the texture, the shape intensity, and temporal data associated with the first scene;
analyze the feature vector using a machine learning model to produce a machine learning model output; and
generate for display to the user profile the recommendation of the second content item based on the machine learning model output.

### Claim 50 (depends on 46)

The system of claim 46, wherein the control circuitry is configured to determine, based on the user interaction, the first content item is of interest to the user profile by:
determining that the first scene of the first content item is of interest to the user profile, based on input received from the user profile.

### Claim 51 (depends on 46)

The system of claim 46, wherein the control circuitry is configured to determine, based on the user interaction, the first content item is of interest to the user profile based on a duration of the user interaction with the first content item and/or a duration of a user interaction with the first scene.

### Claim 52 (depends on 46)

The system of claim 46, wherein the control circuitry is further configured to:
generate for display an icon selectable to cause one or more other scenes similar to the first scene of the first content item, and/or similar to the one or more scenes of the second content item, to be generated for display.

### Claim 53 (depends on 46)

The system of claim 46, wherein the control circuitry is further configured to:
generate for simultaneous display the first scene of the first content item and the recommendation of the second content item, wherein the first scene is displayed during playing of a video corresponding to the first content item to the user profile.

### Claim 54 (depends on 46)

The system of claim 46, wherein the control circuitry is further configured to:
identify a second scene of the first content item;
determine, based on the second scene, a third content item having one or more scenes similar to the second scene; and
generate for display to the user profile a recommendation of the third content item, based on determining that the one or more scenes of the third content item are similar to the second scene.

### Claim 55 (depends on 54)

The system of claim 54, wherein:
the control circuitry is configured to identify the second scene of the first content item by determining that the second scene of the first content item is a current scene being displayed during playing of a video corresponding to the first content item to the user profile, the second scene being later in a duration of the first content item than the first scene; and
the control circuitry is further configured to generate for simultaneous display the second scene of the first content item and the recommendation of the third content item, and cease generating for display the recommendation of the second content item.
