Binary domain generalization for sparsifying binary neural networks

TL;DR

This paper introduces a new binary domain for neural network weights that enhances pruning and compression capabilities of binary neural networks, leading to more efficient models with maintained performance.

Contribution

It proposes a generalized binary domain for BNNs, providing a closed-form quantization solution and demonstrating improved sparsity and efficiency when combined with pruning.

Findings

Achieves reduced memory usage and latency in BNNs.

Maintains performance despite increased sparsity.

Compatible with existing pruning strategies.

Abstract

Binary neural networks (BNNs) are an attractive solution for developing and deploying deep neural network (DNN)-based applications in resource constrained devices. Despite their success, BNNs still suffer from a fixed and limited compression factor that may be explained by the fact that existing pruning methods for full-precision DNNs cannot be directly applied to BNNs. In fact, weight pruning of BNNs leads to performance degradation, which suggests that the standard binarization domain of BNNs is not well adapted for the task. This work proposes a novel more general binary domain that extends the standard binary one that is more robust to pruning techniques, thus guaranteeing improved compression and avoiding severe performance losses. We demonstrate a closed-form solution for quantizing the weights of a full-precision network into the proposed binary domain. Finally, we show the flexibility of our method, which can be combined with other pruning strategies. Experiments over CIFAR-10 and CIFAR-100 demonstrate that the novel approach is able to generate efficient sparse networks with reduced memory usage and run-time latency, while maintaining performance.