Offline Face Detection & Comparison for the Insurance domain using FaceNet | Nitor Infotech
Send me Nitor Infotech's Monthly Blog Newsletter!
×
Software Product Engineering Services Company
  • Company
    • About
    • Leadership
    • Partnership
  • Resource Hub
  • Blog
  • Contact
Software Product Engineering Services Company
Add more content here...
Artificial intelligence Big Data Blockchain and IoT
Business Intelligence Careers Cloud and DevOps
Digital Transformation Healthcare IT Manufacturing
Mobility Product Modernization Software Engineering
Thought Leadership
Aastha Sinha Abhijeet Shah Abhishek Suranglikar
Abhishek Tanwade Abhishek Tiwari Aditya Patil
Ahtesham Shah Ajinkya Pathak Amartya Deshmukh
Amit Pawade Ankita Kulkarni Ankita Patidar
Antara Datta Anup Manekar Ashish Baldota
Chandra Gosetty Deep Shikha Bhat Dr. Girish Shinde
Ekta Shah Ganesh Kadam Gaurav Mishra
Gaurav Rathod Gautam Patil Hariom Ghanshyam Vishwakarma
Harish Singh Chauhan Harshali Chandgadkar Kapil Joshi
Kiran Dhakne Krishna Gunjal Madhavi Pawar
Marappa Reddy Mayur Wankhade Milan Pansuriya
Minal Doiphode Mohit Agarwal Mohit Borse
Nalini Vijayraghavan Nikhil Kulkarni Nikhil Nirawade
Noman Ansari Omkar Ingawale Omkar Kulkarni
Pooja Chavan Pooja Dhule Pranay Gupta
Pranit Gangurde Prashant Kankokar Pratiksha Shelke
Priya Patole Rahul Ganorkar Rashmi Nehete
Ravi Agrawal Robin Pandita Rohan Chavan
Rohini Wwagh Sachin Saini Sadhana Sharma
Sambid Pradhan Sandeep Mali Sanjay Toge
Sanjeev Fadnavis Saurabh Pimpalkar Sayanti Shrivastava
Shardul Gurjar Shishank Sidha Shravani Dhavale
Shreyash Bhoyar Shubham Kamble Shubham Muneshwar
Shubham Navale Shweta Chinchore Sidhant Naveria
Souvik Adhikary Sujay Hamane Tejbahadur Singh
Uddhav Dandale Vasishtha Ingale Vidisha Chirmulay
Vikas Ghadage Yogesh Kulkarni
Artificial intelligence | 05 Jun 2019 |   6 min

Offline Face Detection & Comparison for the Insurance domain using FaceNet

featured image

The insurance industry is estimated as most competitive but a less predictable domain. Insurance policies safeguard against uncertainties and hence are more prone to risks. Therefore, it has always been dependent on statistics. Certainly, the insurance companies have been benefited from data science applications.

One of the most interesting use case we came across is face detection in images and comparison of two or more images to predict whether faces identified in both images are of the same persona.

The detailed requirement is that an insurance executive should be able to capture both images from his/her mobile device. The device will then process those images and compare personas in images to provide outcome.

Despite significant recent advances in the field of face recognition, implementing face verification and recognition efficiently presents challenges to current approaches. Hence, we choose to use the FaceNet system to measure face similarity. Firstly, we have used CNN to identify faces. Secondly, we have used FaceNet to get vector mapping of images and then compare those vector to predict similarity.

What is Facenet?

FaceNet is a system that directly learns a mapping from face images to a compact Euclidean space where distances directly correspond to a measure of face similarity. Once this space has been produced, tasks such as face recognition, verification, and clustering can be easily implemented using standard techniques with FaceNet embeddings as feature vectors.

FaceNet maps a face into a 128D Euclidean space. The L2 distance (or Euclidean norm) between two face embeddings corresponds to its similarity. This is exactly like measuring the distance between two points in a line to know if they are close to each other.

The FaceNet model is a deep convolutional network that employs triplet loss function. Triplet loss function minimizes the distance between a positive and an anchor while maximizing the distance between the anchor and a negative.

Why FaceNet?

FaceNet performs really well on real images. It builds on the Inception ResNet v1 architecture and is trained on the CASIA-WebFace and VGGFace2 datasets. FaceNet’s weights are optimized using the triplet loss function, so that it learns to embed facial images into a 128-dimensional sphere.

A Complete Solution

This is how we ensured a flawless implementation of FaceNet:

  • Accepted two images from user and resized.
  • Detected faces from both the images using Multi-task Cascaded Convolutional Networks (MTCNN).
  • Cropped the faces from the images.
  • Two cropped images were provided as input to face net model, which outputs two vectors of 512 dimensions.
  • Used Tensorflow implementation of MTCNN and FaceNet in android application.
  • Applied Euclidean distance formula to calculated distance between two vectors.
  • Compared distance with predefined threshold value, which can be set as per the analysis result while testing phase.
  • Completed face verification using the result of that comparison.

Screenshots of matches  Screenshot of no matches 

 Conclusion:

Face technology is making its way into the mainstream. Furthermore, there is a growing list of companies offering developer friendly face API through their SaaS platforms, making it a lot easier and cheaper to incorporate state-of-the-art face technology into different types of products.

By now, you should be familiar with how face recognition systems work and how to make your own simplified face recognition system using a version of the FaceNet network.

Related Topics

Artificial intelligence

Big Data

Blockchain and IoT

Business Intelligence

Careers

Cloud and DevOps

Digital Transformation

Healthcare IT

Manufacturing

Mobility

Product Modernization

Software Engineering

Thought Leadership

<< Previous Blog fav Next Blog >>
author image

Nitor Infotech Blog

Nitor Infotech is a leading software product development firm serving ISVs and enterprise customers globally.

   

You may also like

featured image

Demystifying Monorepos: A Code Management Revolution

In the ever-evolving landscape of software development, efficiency, and collaboration stand as the cornerstones of progress. Yet, for many years, engineers and developers found themselves grappling...
Read Blog


featured image

Building and Distributing NPM Packages: A Developer's Guide

In the dynamic world of software development, prioritizing efficiency and collaboration is essential. Enter the npm package, the Node Package Manager, which stands as the largest and most extensive...
Read Blog


featured image

Top 5 Use Cases of NFT: A Closer Look

Our online world allows us to share funny memes, cool pictures, and other interesting visuals with friends and family. Some may do it for fun, others to learn something new, or even to tell people ...
Read Blog


subscribe

Subscribe to our fortnightly newsletter!

We'll keep you in the loop with everything that's trending in the tech world.
Nitor Infotech, an Ascendion company, is an ISV preferred software product development services company. We serve cutting edge GenAI powered services and solutions for the web, Cloud, data, and devices. Our consulting-driven value engineering approach makes it the right fit to be an agile and nimble partner to organizations on the path to digital transformation.

Armed with a digitalization strategy, we build disruptive solutions for businesses through innovative, readily deployable, and customizable accelerators and frameworks.

Company

  • About us
  • Leadership
  • Partnership
  • Press Releases
  • Career
  • Events
  • Contact Us

Quick Links

  • Idea to MVP
  • Product Engineering
  • Platform Engineering
  • Prompt Engineering
  • Blockchain
  • Generative AI

 

  • Big Data & Analytics
  • Mobile App development
  • Cloud and DevOps
  • Internet of Things
  • Blogs
  • Resource Hub

Get in Touch

900 National Pkwy, Suite 210,
Schaumburg, IL 60173,
USA

[email protected]

+1 (224) 265-7110


© 2023 Nitor Infotech All rights reserved

  • Terms of Usage
  • Privacy Policy
  • Cookie Policy
We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it. Accept Cookie policy