Missing Person Face Recognition System using Age invariant Facial Recognition

The system will be a facial recognition system using age invariant facial recognition. It will be able to detect faces even if the person has aged and face has changed. We have implemented different algorithms for image processing. The system will be useful for nonprofit organizations who try to fin

Project Title

Missing Person Face Recognition System using Age invariant Facial Recognition

Project Area of Specialization

Artificial Intelligence

Project Summary

The system will be a facial recognition system using age invariant facial recognition. It will be able to detect faces even if the person has aged and face has changed. We have implemented different algorithms for image processing. The system will be useful for nonprofit organizations who try to find missing people in big cities like Karachi. The system will have a webpage and android app as an interface in which database of missing persons can be entered and viewed

Project Objectives

Our aim is to build a system which can recognize a face even after attributes of the face has changed after years of aging. Generic facial recognition systems have difficulty in this regard. The project aims to recognize face of people regardless of difference in their face over age. To build an application that will assist in maintaining database of missing persons.

Project Implementation Method

We have applied Principal component analysis(PCA), Local Binary Patterns(LBP) and Transfer Learning using Convolutional Neural Networks on the FGnet dataset. It was found that transfer learning produces the best results out of all the three. As this is related to face recognition, we used the pretrained vgg16 model, trained on the VGGFace dataset. We also created our own dataset, by taking pictures from social media sites of our collegues. The FGNet dataset has images of 82 persons, with ages ranging from 1 year to 70 years. Our dataset has images of 21 persons. We used a training-validation split of 0.8/0.2.

Benefits of the Project

The project will benefit Non-government organization which are working to find missing persons. This application that will assist in maintaining database of the missing persons.

Technical Details of Final Deliverable

The project will have a learning model which will be trained on all the images of missing persons. The trained model will be used to classify the missing person if the recent image is tested. The model uses Deep learning model using Transfer learnng. The model is made using python language and libraries like tensorflow and keras. These libraries require good GPU to speed up the learning process of the model. It will have an android app with all the database of the missing persons. The user will be able to add details of the missing person through the app and also find the person in the database by uploading the picture.

Final Deliverable of the Project

Software System

Type of Industry

Others

Technologies

Others

Sustainable Development Goals

Good Health and Well-Being for People, Sustainable Cities and Communities

Required Resources

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