AgeGenderNetwork

By Abhishek Chatterjee(abhishekchatterjeejit@gmail.com)

The aim of this project is to make a android application that detects the gender and age of a person based on the single image of his/her face. This project is using the MobileNetV2 deep learning CNN architecture to predict amd gender and age. The dataset that is used to train is a mix of the IMDB WIKI dataset and Selfie Dataset. The Android application is developed using Java programming langauge. Further OpenCV is also used in several ocations.

Table of Contents

• Introduction
• Dependencies
• File Structure
• Dataset
• Model
• Mobile Application
• Acknowledgments

Introduction

We humans can easily detect the gender of a person by just seeing them, we can probbly detect their age also. It is a very simple task for us.

But what about machines, can they do it?

In this project, I will build a Artificial Intelegence based Android application that can predicts the gender and age of a person just by seeing their face. Further the model will run on the Mobile, not on some powerful server. It will work with live data.

That means the Model needs to be optimized to run on mobile in a very high speed.

Dependencies

There are two parts of the project, one is the Model and the other is the Mobile application.

Model

1. Python 3
2. OpenCV 3
3. Keras

Mobile Application

1. Android Studio
2. OpenCV for Android Studio

File Structure

Dataset

Model

Mobile Application

Acknowledgements