Efficient Secure CNN Inference: A Multi-Server Framework Based on Conditional Separable and Homomorphic Encryption
Longlong Sun, Hui Li, Yanguo Peng, Jiangtao Cui · IEEE Transactions on Cloud Computing · 2024
Deep learning inference has become a fundamental component of cloud service providers, while privacy issues during services have received significant attention. Although many privacy-preserving schemes have been proposed, they require further improvement. In this article, we proposeSerpens, an efficient convolutional neural network (CNN) secure inference framework to protect users’ uploaded data. We introduce a pair of novel concepts, namely separable and conditional separable, to determine whether a layer in CNNs can be computed over multiple servers or not. We demonstrate that linear layers are separable and construct factor-functions to reduce their overhead to nearly zero. For the two nonlinear layers, i.e., ReLU and max pooling, we design four secure protocols based on homomorphic encryption and random masks for two- and n-server settings. These protocols are essentially different from existing schemes, which are primarily based on garbled circuits. In addition, we extensively propose a method to split the image securely. The experimental results demonstrate thatSerpensis$60\times -197\times$faster than the previous scheme in the two-server setting. The superiority ofSerpensis even more significant in the n-server setting, only less than an order of magnitude slower than performing plaintext inference over clouds.