arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.02407cs.CRcs.AI
收藏

HyPHEN:一种用于同态加密神经网络的混合打包方法及优化

HyPHEN: A Hybrid Packing Method and Optimizations for Homomorphic Encryption-Based Neural Networks

Seoul National University(首尔大学)

Donghwan Kim, Jaiyoung Park, Jongmin Kim, Sangpyo Kim, Jung Ho Ahn

AI总结:

针对全同态加密卷积神经网络推理的高开销问题,提出HyPHEN混合打包方法,结合新卷积算法与优化,将CIFAR-10推理降至1.4秒并首次实现ImageNet推理。

摘要:

使用全同态加密(FHE)的卷积神经网络(CNN)推理是一种有前景的私有推理(PI)解决方案,因为FHE能够将整个计算过程外包给服务器,同时保护敏感用户数据的隐私。先前基于FHE的CNN(HCNN)工作已证明了使用FHE构建如ResNet等深度神经网络架构的可行性。尽管取得了这些进展,但由于高昂的计算和内存开销,HCNN在实际应用中仍面临重大挑战。为了克服这些限制,我们提出了HyPHEN,一种深度HCNN构建方法,它结合了新颖的卷积算法(RAConv和CAConv)、数据打包方法(2D间隙打包和PRCR方案)以及针对HCNN构建定制的优化技术。这些增强使得HyPHEN能够大幅减少内存占用以及昂贵的同态操作(如密文旋转和自举)的数量。因此,HyPHEN将HCNN在CIFAR-10上的推理延迟降低到实际可用的1.4秒(ResNet-20),并首次展示了HCNN在ImageNet上的推理,延迟为14.7秒(ResNet-18)。

英文摘要:

Convolutional neural network (CNN) inference using fully homomorphic encryption (FHE) is a promising private inference (PI) solution due to the capability of FHE that enables offloading the whole computation process to the server while protecting the privacy of sensitive user data. Prior FHE-based CNN (HCNN) work has demonstrated the feasibility of constructing deep neural network architectures such as ResNet using FHE. Despite these advancements, HCNN still faces significant challenges in practicality due to the high computational and memory overhead. To overcome these limitations, we present HyPHEN, a deep HCNN construction that incorporates novel convolution algorithms (RAConv and CAConv), data packing methods (2D gap packing and PRCR scheme), and optimization techniques tailored to HCNN construction. Such enhancements enable HyPHEN to substantially reduce the memory footprint and the number of expensive homomorphic operations, such as ciphertext rotation and bootstrapping. As a result, HyPHEN brings the latency of HCNN CIFAR-10 inference down to a practical level at 1.4 seconds (ResNet-20) and demonstrates HCNN ImageNet inference for the first time at 14.7 seconds (ResNet-18).

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑