# US 2019/0378171 A1 · Targeted advertisement system

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

- Number: US 2019/0378171 A1
- Document ID: US20190378171A1
- Status: Published application
- Kind: A1
- Application number: 16/433,771
- Filed: 2019-06-06
- Published: 2019-12-12
- Earliest claimed: 2018-06-06
- Assignees: Walmart Apollo, LLC
- Applicants: Walmart Apollo, LLC
- Inventors: Rajesh Shreedhar Bhat; Souraj Mishra; Kushal M. Kumar; Abhishek Ramesh; Nanda Kishore Reddy
- Web page: https://patentgrep.com/patent/US20190378171A1

## CPC

- G06Q 30/0269: {based on user profile or attribute}
- G06F 18/2411: based on the proximity to a decision surface, e.g. support vector machines
- G06F 18/2413: based on distances to training or reference patterns
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06V 10/454: {Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]}
- G06V 10/764: using classification, e.g. of video objects
- G06V 40/161: {Detection; Localisation; Normalisation}
- G06N 3/045: Combinations of networks
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- G06V 40/178: {estimating age from face image; using age information for improving recognition}
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Systems, methods, and machine readable media are provided for targeted advertising. One or more images of individuals in a location in a facility are captured with an imaging device. Face detection is performed on the one or more images. Based on a face detection result, Convolutional Neural Network models are used to determine a gender and estimate an age of each detected face. A targeted audience is determined, based on pre-determined criteria using the age estimation and gender identification of each detected face. A relevant advertisement is selected for the targeted audience for display on a display device. The relevant advertisement may be selected based at least in part on a footfall score (fscore) indicative of customer traffic in the facility.

## Claims

### Claim 1

A system for targeted advertising, the system comprising:
an imaging device configured to take one or more images of a plurality of individuals at a location in a facility, wherein the one or more images are at least one of still and video images;
a display device configured to display an advertisement at the location;
a computing device in communication with the imaging device and the display device and configured to execute a targeted advertising module, wherein the targeted advertisement module, when executed:
receives the one or more images from the imaging device;
performs face detection on the one or more images;
produces, based on the faces detected, age estimation and gender identification for each detected face using a Convolutional Neural Network (CNN) based deep learning model;
determines a targeted audience based on pre-determined criteria using the age and gender identification of each detected face; and
selects a relevant advertisement for the targeted audience for display on the display device, wherein the relevant advertisement is selected based at least in part on a footfall score (fscore) indicative of customer traffic in the facility.

### Claim 2 (depends on 1)

The system of claim 1, wherein the computing device uses a linear support vector machine (SVM) to perform face detection on the one or more images.

### Claim 3 (depends on 1)

The system of claim 1, wherein the targeted advertising module divides the at least two detected faces into one of a group comprising teenage, adult, middle age adult, or older aged adult based on the age identification.

### Claim 4 (depends on 1)

The system of claim 1, wherein the targeted advertising module displays a relevant advertisement based on pre-categorization of products based on their pre-defined appropriateness in regards to gender and age.

### Claim 5 (depends on 1)

The system of claim 1, wherein the footfall score (fscore) is generated based on mapping of a product category with a number of customers, a location within the store, and a score proportional to footfall.

### Claim 6 (depends on 1)

The system of claim 1, wherein the product category score is determined according to the formula
Scorei=αpscore+(1−α)ascore,
and is used to determine the relevant advertisement.

### Claim 7 (depends on 6)

The system of claim 6, wherein the product score (pscore) is generated according to the formula
f
score
-
min

(
f
score
)
max

(
f
score
)
-
min

(
f
score
)

### Claim 8 (depends on 6)

The system of claim 6, wherein an advertisement score (ascore) is generated that is proportional to a number of advertisements relevant to a category of items.

### Claim 9

A computer-implemented method for targeted advertising, the method comprising:
capturing one or more images of a plurality of individuals in a location in a facility with an imaging device, wherein the one or more images are at least one of still and video images;
performing face detection on the one or more images;
producing, based on the faces detected, age estimation and gender identification for each detected face using a Convolutional Neural Network (CNN) based deep learning model;
determining a targeted audience based on pre-determined criteria using the age and gender identification of each detected face; and
selecting a relevant advertisement for the targeted audience for display on a display device, wherein the relevant advertisement is selected based at least in part on a footfall score (fscore) indicative of customer traffic in the facility.

### Claim 10 (depends on 9)

The computer-implemented method of claim 9, wherein performing face detection on the one or more images includes using a linear support vector machine (SVM).

### Claim 11 (depends on 9)

The computer-implemented method of claim 9, wherein the determining a targeted audience includes dividing the at least two detected faces into one of a group comprising teenage, adult, middle age adult, or older aged adult based on the age identification.

### Claim 12 (depends on 9)

The computer-implemented method of claim 9, wherein displaying a relevant advertisement by the targeted advertising module includes pre-categorization of products based on their pre-defined appropriateness in regards to gender and age.

### Claim 13 (depends on 9)

The computer-implemented method of claim 9, further comprising generating the footfall score (fscore) based on mapping of a product category with a number of customers, a location within the store, and a score proportional to footfall.

### Claim 14 (depends on 9)

The computer-implemented method of claim 9, wherein the product category score is determined according to the formula
Scorei=αpscore+(1−α)ascore,
and is used to determine the relevant advertisement.

### Claim 15 (depends on 14)

The computer-implemented method of claim 14, wherein the product score (pscore) is generated according to the formula
f
score
-
min

(
f
score
)
max

(
f
score
)
-
min

(
f
score
)

### Claim 16 (depends on 14)

The computer-implemented method of claim 14, wherein an advertisement score (ascore) is generated that is proportional to a number of advertisements relevant to a category of items.

### Claim 17

A non-transitory machine-readable medium storing instructions executable by a computing device, wherein execution of the instructions causes the computing device to implement a method for targeted advertising, the method comprising:
capturing one or more images of a plurality of individuals in a location in a facility with an imaging device, wherein the one or more images are at least one of still and video images;
performing face detection on the one or more images;
producing, based on the faces detected, age estimation and gender identification for each detected face using a Convolutional Neural Network (CNN) based deep learning model;
determining a targeted audience based on pre-determined criteria using the age and gender identification of each detected face; and
selecting a relevant advertisement for the targeted audience for display on a display device, wherein the relevant advertisement is selected based at least in part on a footfall score (fscore) indicative of customer traffic in the facility.

### Claim 18 (depends on 17)

The non-transitory machine-readable medium of claim 17, wherein the determining a targeted audience includes dividing the at least two detected faces into one of a group comprising teenage, adult, middle age adult, or older aged adult based on the age identification.

### Claim 19 (depends on 17)

The non-transitory machine-readable medium of claim 17, wherein displaying a relevant advertisement by the targeted advertising module includes pre-categorization of products based on their pre-defined appropriateness in regards to gender and age.

### Claim 20 (depends on 19)

The non-transitory machine-readable medium of claim 19, wherein the footfall score is generated based on mapping of a product category with a number of customers, a location within of the store, and a score proportional to footfall; wherein a product score (pscore) is generated according to the formula
f
score
-
min

(
f
score
)
max

(
f
score
)
-
min

(
f
score
)
;
wherein an advertisement score (ascore) is proportional to a number of advertisements relevant to a category of items; and wherein the product category score (Scorei) is determined according to the formula
Scorei=αpscore+(1−α)ascore.
