提出专用于二值化网络的剪枝框架与全局加权算法，显著提升压缩率。
Pruning Binarized Neural Networks: A Dedicated Framework and Globally Weighted Algorithms
- 设计专用框架支持二值化网络的冻结与剪枝
- 在VGG11上实现70%剪枝率且精度不变
- 适合边缘设备部署的高效模型优化
深度神经网络的极端压缩，包括完全二值化，能大幅降低内存占用和计算复杂度，有利于在资源受限的边缘硬件（如FPGA和微控制器）上部署。尽管将二值化与剪枝结合可进一步提升效率，但现有剪枝策略不适用于二值化表示，且极少带来实际硬件收益。本文提出一个基于PyTorch的研究型框架，集成冻结与剪枝机制，用于设计和优化二值化神经网络。该框架支持对前沿方法的快速复现与新方法的敏捷原型开发。基于此框架，我们提出一种新型剪枝方法，考虑不同抽象层级下参数的相对重要性。该全局加权机制在模型准确率与剪枝率之间实现了更优平衡，在VGG11上达到70%剪枝率且精度保持不变，而现有最优结果在二值化设置下仅为41%。
原文摘要 · Abstract (English)
Extreme compression of deep neural networks, up to full binarization, dramatically reduces memory footprint and arithmetic complexity, facilitating deployment on constrained edge hardware with field-programmable gate arrays (FPGAs) and microcontrollers. Although combining binarization with pruning promises additional efficiency gains, existing pruning strategies are ill-suited to binarized representations and rarely translate into meaningful hardware savings. We introduce a PyTorch-based, research-oriented framework that incorporates freezing and pruning mechanisms for designing and optimizing binarized neural networks. The framework enables rapid and reproducible evaluation of state-of-the-art approaches and the fast prototyping of new ones. Leveraging this framework, we propose a novel pruning method that accounts for the relative importance of learned parameters across abstraction levels. Such a global weighting mechanism consistently achieves a superior trade-off between model accuracy and pruning rate, achieving a 70% pruning rate on VGG11 with constant accuracy, while state-of-the-art results reach only 41% in the binarized setting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。