Sightcorp logo

Facial Recognition Software Companies

Everything about Facial Recognition Software Companies

     

Facial Recognition Software Companies

Facial recognition is a fast-growing field, and besides Sightcorp, there are numerous other software companies working on this technology. Some of the biggest names include:

  • Amazon
  • Face++
  • Microsoft

 

What do facial recognition software companies do?

These companies develop AI-based software that can detect faces in images and video files and subsequently verify and identify faces for a wide range of applications. Amazon’s Rekognition, for example, can be used in situations varying all the way from law enforcement to celebrity recognition and image search.

When it comes to facial recognition, software companies other than tech giants like Amazon and Microsoft, usually specialize in a narrower field of application.

At Sightcorp, for example, our FaceMatch technology aims to meet the specific needs of customers who are looking for solutions in the following areas:

  • Know Your Customer (KYC): For banks and other financial service providers who need to verify and authenticate their customers and reduce identity fraud.
  • Access Control: For securing access to buildings, events, devices, and more.
  • Payment Verification: For adding an extra layer of security to payment transactions, improving the user experience during payment transactions, and reducing fraud.

 

Click here to learn more about FaceMatch

There are also companies that specialize in facial recognition software for:

  • Law enforcement and retail security: To carry out surveillance and recognize suspects and known criminals.
  • Automotive: To enhance the driving experience, for example by setting up different profiles for each authorized driver and adjusting the car’s settings depending on who gets into the driver’s seat.
  • Healthcare: To improve the patients’ experience and to ensure that the right care is given to the right patient at the right time. 

What to consider when comparing facial recognition software companies

Accuracy rates: How accurate is the company’s face recognition software? And are you able to customize the confidence thresholds to suit your particular use case?

Sightcorp’s software performs with state-of-the-art accuracy.

Click here to learn more.

Underlying technology: What is the underlying technology that the company uses in their facial recognition software? Are they still using traditional AI techniques, or are they using newer deep learning methods?

Sightcorp’s FaceMatch is 100% deep learning-based.

Area of specialization: Does the company provide a solution for your use case? Is it one of their focus areas? If not, do they have the resources and expertise to develop a solution that is tailored to your needs?

At Sightcorp, we specialize in KYC, Access Control, and Payment Verification. If you have a different use case in mind, however, please feel free to reach out to us so that we can discuss the possibilities.

Country: Does the company operate in – and understand the local circumstances of – the country or countries where you want to implement facial recognition?

Sightcorp operates globally, and we pay close attention to regional differences in the context of facial recognition.

Data security: Do you have a choice of where your data is stored and processed? Generally speaking, it is more secure to store your data locally, where you have full control over it. In practice, though, this may not always be the most suitable option based on your specific needs.

With Sightcorp’s FaceMatch SDK, your data is stored and processed locally. With FaceMatch API (coming soon), you will be able to store and process your data in the cloud. The choice is yours!

Privacy: How does the company interact with privacy laws?

At Sightcorp, we do not store any of the facial images processed by FaceMatch. The software only stores the face embeddings, which on their own cannot be traced back to any individual.

Click here to learn more about Sightcorp’s FaceMatch

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.