用同态加密保护模型推理隐私，实现高精度低延迟的加密计算。
PRISM: Privacy-preserving Inference System with Homomorphic Encryption and Modular Activation
- 用多项式近似替代非线性激活函数，适配同态加密
- CIFAR-10上达94.4%准确率，单样本加密推理仅2.42秒
- 适合需隐私保护的医疗、金融等敏感领域部署
随着机器学习快速发展，模型在各行业具备强大预测能力。然而，在关键基础设施中部署时，数据隐私顾虑限制了数据共享。同态加密（HE）可在加密数据上直接计算，但难以兼容依赖非线性激活函数的卷积神经网络（CNN）。为此，本文提出优化框架，将标准非线性函数替换为同态兼容的近似形式，兼顾安全计算与低开销。通过重构CNN结构并引入高效激活函数近似方法，缓解加密带来的性能损失。在CIFAR-10上的实验表明，采用度为4的多项式与Softplus激活函数配合CKKS方案，单个加密样本推理耗时2.42秒，10,000个样本共需24,000秒，准确率达94.4%，在隐私与效率间取得良好平衡。
原文摘要 · Abstract (English)
With the rapid advancements in machine learning, models have become increasingly capable of learning and making predictions in various industries. However, deploying these models in critical infrastructures presents a major challenge, as concerns about data privacy prevent unrestricted data sharing. Homomorphic encryption (HE) offers a solution by enabling computations on encrypted data, but it remains incompatible with machine learning models like convolutional neural networks (CNNs), due to their reliance on non-linear activation functions. To bridge this gap, this work proposes an optimized framework that replaces standard non-linear functions with homomorphically compatible approximations, ensuring secure computations while minimizing computational overhead. The proposed approach restructures the CNN architecture and introduces an efficient activation function approximation method to mitigate the performance trade-offs introduced by encryption. Experiments on CIFAR-10 achieve 94.4% accuracy with 2.42 s per single encrypted sample and 24,000 s per 10,000 encrypted samples, using a degree-4 polynomial and Softplus activation under CKKS, balancing accuracy and privacy.
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。