XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks

TL;DR

This paper introduces XNOR-Networks, a highly efficient binary convolutional neural network architecture that significantly reduces memory usage and increases speed, enabling real-time ImageNet classification on CPUs with minimal accuracy loss.

Contribution

The paper presents a novel binary convolutional network approach that outperforms previous binarization methods in accuracy and efficiency, enabling practical deployment on standard hardware.

Findings

58x faster convolutional operations

32x memory savings

2.9% accuracy loss on ImageNet compared to full-precision AlexNet

Abstract

We propose two efficient approximations to standard convolutional neural networks: Binary-Weight-Networks and XNOR-Networks. In Binary-Weight-Networks, the filters are approximated with binary values resulting in 32x memory saving. In XNOR-Networks, both the filters and the input to convolutional layers are binary. XNOR-Networks approximate convolutions using primarily binary operations. This results in 58x faster convolutional operations and 32x memory savings. XNOR-Nets offer the possibility of running state-of-the-art networks on CPUs (rather than GPUs) in real-time. Our binary networks are simple, accurate, efficient, and work on challenging visual tasks. We evaluate our approach on the ImageNet classification task. The classification accuracy with a Binary-Weight-Network version of AlexNet is only 2.9% less than the full-precision AlexNet (in top-1 measure). We compare our method with recent network binarization methods, BinaryConnect and BinaryNets, and outperform these methods by large margins on ImageNet, more than 16% in top-1 accuracy.