Effective Tracking of Passing People for Marketing using Deep Learning

Today Pakistan?s local brands need to approach for new marketing schemes in order to attract new customers as well as to retain the existing ones so they can compete with their foreign competitors. Many retailing companies are gathering data of their customer?s on basis of age and gender but it is a

Project Title

Effective Tracking of Passing People for Marketing using Deep Learning

Project Area of Specialization

Artificial Intelligence

Project Summary

Today Pakistan’s local brands need to approach for new marketing schemes in order to attract new customers as well as to retain the existing ones so they can compete with their foreign competitors. Many retailing companies are gathering data of their customer’s on basis of age and gender but it is a manual process to do it. The solution that we are providing is basically based on hyper-targeting which is focused on individual group of people that will help retailers/marketers to have a complete knowledge of their customer’s demographics. The primary aim of this project is to determine in-store activity, help marketers analyze performance and success of their new product, manage staff schedules according to peak periods and maximize the sales potential. We are doing gender and age group segmentation with the help of deep learning algorithm i.e. Mini-Xception. In order to analyze the trends given from the statistical information through our system, customer’s information is displayed on a dynamic web application. The end application is a web service on which accurate demographics of the customers will be shown.

Project Objectives

Marketing strategies cannot be formed without knowing your audience towards a specific product or brand. In today’s world of marketing customers demographics play very important role. To help marketers analyze a complete record of their customer’s demographics we are proposing a solution which will accurately identify the age group and gender of the consumers.

Project Implementation Method

The very basic idea of the system is to classify gender and age group in real time.  First live feed is being obtained from an IP Camera. This live feed is then given to the system in the form of frames. For the extraction of human faces from the incoming frames, Dlib face detection is being used. The extracted faces are the region of interest and are being used for testing the results of our system. The faces are passed to the convolutional neural network, which have been trained using the IMDB-WIKI Dataset, to classify them into their respective gender and age groups. Each time a certain gender or age group are being detected; they are also being counted. These results are stored in a database and displayed using a web application that is made available to the user.

Benefits of the Project

This system will help shop owner’s compare the amount of transactions with the people visiting
the outlet.

Before opening an outlet in a new area, the system can be installed to check whether the core
demographic visits that region.

This system will help reduce manual labour.

Will record peak periods of customers, so that staffing pattern can be matched with customer
flow.

Helps maximize sales potential and to analyze success of marketing initiatives.

Technical Details of Final Deliverable

The hardware portion of this system consists of an IP Camera.  The Camera is simply being used to obtain live feed and being fed into the system in the form of frames for gender and age classification. The software portion of the system is a web application which displays the data collected

Final Deliverable of the Project

HW/SW integrated system

Core Industry

IT

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Others

Sustainable Development Goals

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com