Sightcorp logo

Face Matching Software

Everything about Face Matching Software

     

What is Face Matching software?

Face matching software uses AI algorithms to determine whether two facial images are of the same person. This software can be used by:

  • Banks and other financial service providers to carry out Know Your Customer (KYC) compliance procedures and payment verifications
  • Sport, music, and other entertainment venues to facilitate access control and enhance security
  • Airports and border authorities to streamline access control and more effectively identify potential threats
  • Hotels to facilitate and automate the check-in process
  • Retail stores to flag known shoplifters, identify VIP shoppers, and introduce frictionless shopping
  • Smartphone and other consumer electronics manufacturers to improve device security
  • Healthcare providers to ensure that the correct patients receive the correct treatments

 

As you can see from the above use cases, face matching software is often integrated with a broader system to achieve a particular goal related to face identification and/or verification.

Sightcorp’s face matching software

At Sightcorp, our face matching software is proprietary and deep learning-based. We currently offer FaceMatch SDK, which you can use to build your own facial recognition solution for Windows and Linux platforms. In order to further develop with our SDK into new applications, C++ expertise is required.

The face matching process starts with converting images into individual pixels and running the algorithm across the pixels to detect a face in the image. If a face is detected, the algorithms subsequently convert the face into a faceprint or face embedding, and then compare these faceprints or face embeddings to determine how similar they are. If the faceprints or embeddings are sufficiently similar based on the selected confidence threshold (which can be customized according to your own needs), the software indicates that it is confident that the faces belong to the same person and there is a match.

With FaceMatch SDK, image processing takes place locally after a one-time license authentication connection, which means that you have full control over your data security.

For even more customized solutions, we are also able to integrate our pre-trained face detection and recognition models into your processes as “customizable building blocks”.

In the near future, we will also be offering FaceMatch as a web-based API service, enabling you to integrate facial recognition into your Cloud-based systems and applications.

Learn More About FaceMatch

How to choose the right face matching software for your needs

The various software solutions on the market each have their own specialized area of focus. While a particular solution might be the best option in one industry, a different solution may be the better option in another, for example.

To help you figure out which aspects to compare and consider, ask yourself the following questions:

  • What is the main purpose for which I want to use the software?
  • Do I want to recognize or compare faces in images, videos, or both?
  • Am I looking for a ready-to-use, standalone product, or do I want to develop something that can be integrated into existing applications?
  • How customizable do I want the software to be? Do I want to set the confidence thresholds myself? Do I want to re-train the models on my own or my customers’ data?
  • How many images, on average, will I be processing each month?
  • Which platform(s) do I want to run the software on? Do I want to process locally or in the Cloud?

 

Once you’ve answered these questions, you’ll be in a good position to compare the various face matching software solutions that are available to you. 

Discover Sightcorp’s FaceMatch Solution

Here are other articles you might find interesting:

 

   

Technical Specifications

The table below shows how FaceMatch SDK performs on the Labelled Faces in the Wild (LFW) dataset:

FPRTPRThreshold (Inverse of distance)
0.10.99900 ±0.002130.55448
0.010.99667 ±0.005370.59791
0.0010.99367 ±0.006050.62989

FPR = False Positive Rate
TPR = True Positive Rate

These results are an indication only and are based on the specific dataset Labelled Faces in the Wild. Customers can expect similar performance, with possible variations due to hardware and the availability of annotated data.