Implementation of selected neural network architectures for biometric face identification and their functional tests

Date: 2019
Customer: Inwebit
Project Manager: Marek Kraft

The aim of the project was to implement selected classifiers in the form of convolution neural networks. The task of the classifiers was biometric recognition of people based on facial images. The embedding vector generation approach was used and an accuracy of over 99.7% was obtained on the Labeled Faces in the Wild database with 1680 identities.