Rectified Binary Convolutional Networks with Generative Adversarial Learning

Binarized convolutional neural networks (BNNs) are widely used to improve the memory and computational efficiency of deep convolutional neural networks for to be employed on embedded devices. However, existing BNNs fail to explore their corresponding full-precision models’ potential, resulting in a significant performance gap. This paper introduces a Rectified Binary Convolutional Network (RBCN) by combining full precision kernels and feature maps to rectify the binarization process in a generative adversarial network (GAN) framework. We further prune our RBCNs using the GAN framework to increase the model efficiency and promote flexibly in practical applications. Extensive experiments validate the superior performance of the proposed RBCN over state-of-the-art BNNs on tasks such as object classification, object tracking, face recognition, and person re-identification.