Facial recognition – New era for online identification
Everyone has already tested in its life a facial recognition system such as those installed in airports. Relying on the reading of NFC chips, this type of solution enables the extraction of the picture from the ID and compare it with a selfie of the user. Very reliable, these technologies require however a very high level of quality of the picture that can be achieved because the camera is part of the system.
When dealing with online identification and facial recognition, the challenge is that the reading of the picture on the NFC chip is most of the time not available, that the camera is not managed by the provider and can be a very low quality sensor, …. However, the needs of comparing the end-user behind its camera with the picture on the ID remains valid.
Facial recognition at AriadNEXT
At AriadNEXT, we’ve been working with Scientists to develop a new generation of facial recognition system based on Deep Learning algorithm. Provided as a complement of our ID Verification Services, this new service enables the comparison of the cropped picture from the ID with a Selfie of the end-user. Either its matches or it doesn’t.
We already have impressing results when testing public database such as the LWF1.
This service will be included in our Identification Services as well as our Digital Identity offering.
Coming soon, we will enhance our SDK to provide our Clients with Liveness detection features in our SDK.
1The Labeled Wikipedia Faces (LWF) is a dataset of 8,500 faces for about 1,500 identities, taken from Wikipedia. The LWF facial images are aligned with faces in the Labeled Faces in the Wild database so face verification experiments can be performed and compared.