InsightFace LARQ Keras code

My ultra-efficient approach for Very Low Resolution Face Recognition using Binary Neural Networks (BNN), named BinaryFaceNet, is designed for maximizing efficiency performance, while having competent face identification and verification performance on state-of-the-art benchmark datasets for the Very Low Resolution Face Recognition problem, compared to state-of-the-art BNNs. This is one of the outcomes of my PhD thesis project. The code was developed on Python using Keras over TensorFlow and the LARQ framework.

See more at:
Link to Repository
Link to Thesis