CHE: Channel-Wise Homomorphic Encryption for Ciphertext Inference in Convolutional Neural Network

Abstract

Privacy-preserving deep learning (PPDL), which leverages the Homomorphic Encryption (HE), has attracted attention as a promising approach to ensure the privacy of deep learning applications' data. While recent studies have developed and evaluated the HE-based PPDL algorithms, the achieved performances, such as accuracy and latency, need improvement to make the applications practical. This work aims to improve the performance of the image classification of HE-based PPDL by combining two approaches -Channel-wise Homomorphic Encryption (CHE) and Batch Normalization (BN) with coefficient merging. Although these are commonly used schemes, their detailed algorithms and formulations have not been clearly described.The main contribution of current study is to provide complete and reproducible descriptions of these schemes. We evaluate our CHE and BN implementation by targeting the Cheon-Kim-Kim-Song scheme as an HE scheme and Convolution Neural Network (CNN) as a machine learning scheme while using the MNIST and CIFAR-10 as the datasets. In addition, we compare the results with the five state-of-the-art neural network architectures. Our experiments demonstrate that the CHE can serve as a tool for empirically achieving shorter latency (the shortest 7.76 seconds) and higher accuracy (the highest 99.32%) compared with the previous studies that aimed to establish the classification of the encrypted MNIST data with CNN. Our approach can aid in designing a more robust and flexible PPDL.

Efficient and Privacy-Preserving Image Classification Using Homomorphic Encryption and Chunk-based Convolutional Neural Network

“…Their innovative approach combined Channel-wise Homomorphic Encryption (CHE) with Batch Normalization (BN) and coefficient merging. Their comprehensive evaluations using the Cheon-Kim-Kim-Song HE scheme on datasets like MNIST and CIFAR-10 revealed significant improvements in both accuracy and latency [5].…”

“…If the gradient value of a split chunk exceeds this threshold, it is considered that the split chunk contains feature information. This criterion helps us distinguish between split chunks with important features and those without, allowing us to focus our computational resources on fragments that are more likely to contribute to the overall image classification [5].…”

Efficient and privacy-preserving image classification using homomorphic encryption and chunk-based convolutional neural network

“…If a segment's gradient value surpasses this threshold, it is considered to contain ample feature information; otherwise, it is deemed to lack critical feature information. This criterion aids us in distinguishing segments with significant features from those without, allowing us to concentrate computational resources on segments that are more likely to contribute to the overall image classification [38]. The entire chunking and gradient computation process above is shown in Fig.…”

PAF-FHE: Low-Cost Accurate Non-Polynomial Operator Polynomial Approximation in Fully Homomorphic Encryption Based ML Inference

“…1) with our contributions as follows. [15], CryptoGCN [20] ✗ ✗ ✓ CryptoNet [5], CryptoDL [7], LoLa [13], CHE [26] ✗ ✗ ✓ F1 [24], CraterLake [25], BTS [10] ✓ ✓ ✗ HEAX [22],Delphi [17],Gazelle [8],Cheetah [21] ✗ ✗ ✓ SHE [14] ✓ 2 Technical Background…”

“…Moreover, the continuous streaming data in cloud applications also invalidate the Multi-Party Computation [23] and Garbled Circuits (GC) [9] because both of them rely on prohibitively high communication overhead. To eliminate the overhead of communication as well as the secure data sharing, non-polynomial operators can be approximated by low-degree polynomials for inference [2,5,7,13,18,26]. In the current FHE schemes which most prior arts focus on, non-polynomial operators like ReLU or MaxPooling are not supported.…”