Face Detection API is just one of Microsofts Cognitive Services that works with peoples images, specifically with their faces.

The ¨Face Detection API¨ detects up to 64 faces with high precision and locates them in an image.

To do this, it provides a rectangle indicating the position of the face in the image as well as a range of precise data, for example to indicate whether it is a man or a woman, if they are wearing glasses, have a mustache, or even if they are smiling.

Face Detection API functions

It provides four Facial Recognition areas:

Face verification:

The face verification API performs an authentication against two detected faces, in order to know if they belong to the same person. The API returns a confidence score on how probable it is that the two faces belong to the same person.


Finding similar faces:

Given the identified face and a set of candidate faces to search through, the service finds the group of faces most similar to the target face.

There are two operating modes in order to find similar faces:

MatchPerson: Returns similar faces after applying face verification.
MatchFace: Ignores face verification and returns faces similar to the original.

In this example we can see a list of candidates faces: To find four similar faces, MatchPerson returns faces (a) and (b) that belong to the same person. MatchFace on the other hand returns faces (a), (b), (c) and (d), even if they have no similarity.

And the detected face is:

 

Grouping of faces:

Face Detection API can also group together the faces of several people by similarity. For example, it is able to know that several photos are of the same person without knowing this previously.

Identification of faces:
The Faces API can be used to identify people based on a detected face and a database of people (person group), which must of been created previously.

The image below is an example of a person group called “Myfriends”. Each group can contain up to 1000 individuals (In the example, Anna, Bill and Clare are the people), and each person can have one or more faces registered (As we see in the example, Anna has several registered faces).

After a group has been created and trained, identification can be made against the group and the new face detected.
If the face is identified as an individual already in the group, the system will return that person. (It will tell us that the face detected belongs to Anna).

Face Storage:
Face storage allows a standard subscription to store additional persistent faces when using people or face lists for face identification or face search with the faces API.

The stored images cost $0.5 for every 1000 faces and this ratio is prorated daily. Free subscriptions are limited to a total of 1000 people.

Applicability to business: 
To which business sectors can we apply the Facial Recognition?

The possibilities of using Facial Recognition in business applications are unlimited. We can make an app recognize people by just showing an image or a video in streaming.

This is especially interesting in fields such as Security: In airports, access controls, user authentication, etc. Facial Recognition is practically essential.

On the other hand, due to the large amount of data provided by the user through a single photograph, we can customize, and greatly improve the offers we present to our customers, and thus increase sales of our products.

Isn’t it much easier to construct and communicate an offer if we understand our clients and what they like? And in addition to being simpler, it’s also much more efficient!

In short, any sector that works with people and end users (B2C), can benefit from Facial Recognition thanks to the large amount of information that it provides us with about our clients, so what are you waiting for? Get started!

.