With the rapid progress of face recognition it has more and more applications in everyday life. Although its backbone,very deep neural networks, also show improvement both in terms of accuracy and efficiency their computational cost andmemory usage is still a limiting factor for deploying these models on a hardware with limited computational and powerresources, such as mobile or embedded devices. Here arises the task of learning fast and compact deep neural networkswhich have a comparable accuracy to the complex model as requirement of real-life applications. Another issue is thatsometimes face recognition system may run models of different complexity depending of the devices used for biometrictemplate extraction (i.e. desktop with GPU or mobile phone), so the compatibility between the face descriptors isdesirable. Our paper considers both this cases: we propose a new method for learning fast and compact face recognitionmodel which has a similar performance to a much more complex model used for transferring its knowledge and we alsoshow that both these models can be used for verification in a single face recognition system. To the best of ourknowledge such evaluation of a compatibility between 2 different models for face recognition was never done before ourwork.
展开▼