Research and develop a deep learning-based tool for face recognition.

Abstract

In the real world, the Deep convolutional neural network[18,19,20,21] (DCNNs) model for face recognition tasks, including face verification and face identification has been applied in many perspectives of life. One of the main challenges of face recognition models is the design of appropriate loss functions that enhance discriminative power. One of the key factors in supporting the high-quality model is the loss function which plays an important role in achieving high performance of feature learning.