RBCN: Rectified Binary Convolutional Networks for Enhancing the Performance of 1-bit DCNNs

TL;DR

This paper introduces RBCNs, a novel approach that combines full-precision and binary components using GANs to significantly improve the performance of 1-bit DCNNs, especially in resource-constrained applications.

Contribution

The paper proposes a unified framework for rectified binary convolutional networks that leverages full-precision guidance via GANs to enhance binarized neural network performance.

Findings

RBCNs outperform state-of-the-art BCNNs in accuracy.

The method generalizes well to object tracking tasks.

Extensive experiments validate the effectiveness of RBCNs.

Abstract

Binarized convolutional neural networks (BCNNs) are widely used to improve memory and computation efficiency of deep convolutional neural networks (DCNNs) for mobile and AI chips based applications. However, current BCNNs are not able to fully explore their corresponding full-precision models, causing a significant performance gap between them. In this paper, we propose rectified binary convolutional networks (RBCNs), towards optimized BCNNs, by combining full-precision kernels and feature maps to rectify the binarization process in a unified framework. In particular, we use a GAN to train the 1-bit binary network with the guidance of its corresponding full-precision model, which significantly improves the performance of BCNNs. The rectified convolutional layers are generic and flexible, and can be easily incorporated into existing DCNNs such as WideResNets and ResNets. Extensive experiments demonstrate the superior performance of the proposed RBCNs over state-of-the-art BCNNs. In particular, our method shows strong generalization on the object tracking task.