Lightening Encrypted Convolutions: A GPU-Optimized Approach to Private Inference

COVID-19 is highly infectious and causes acute respiratory disease. Machine learning (ML) and deep learning (DL) models are vital in detecting disease from computerized chest tomography (CT) scans. The DL models outperformed the ML models. For COVID-19 detection from CT scan images, DL models are used as end-to-end models. Thus, the performance of the model is evaluated for the quality of the extracted feature and classification accuracy. There are four contributions included in this work. First, this research is motivated by studying the quality of the extracted feature from the DL by feeding these extracted to an ML model. In other words, we proposed comparing the end-to-end DL model performance against the approach of using DL for feature extraction and ML for the classification of COVID-19 CT scan images. Second, we proposed studying the effect of fusing extracted features from image descriptors, e.g., Scale-Invariant Feature Transform (SIFT), with extracted features from DL models. Third, we proposed a new Convolutional Neural Network (CNN) to be trained from scratch and then compared to the deep transfer learning on the same classification problem. Finally, we studied the performance gap between classic ML models against ensemble learning models. The proposed framework is evaluated using a CT dataset, where the obtained results are evaluated using five different metrics The obtained results revealed that using the proposed CNN model is better than using the well-known DL model for the purpose of feature extraction. Moreover, using a DL model for feature extraction and an ML model for the classification task achieved better results in comparison to using an end-to-end DL model for detecting COVID-19 CT scan images. Of note, the accuracy rate of the former method improved by using ensemble learning models instead of the classic ML models. The proposed method achieved the best accuracy rate of 99.39%.

Although the existing traditional image classification methods have been widely applied in practical problems, there are some problems in the application process, such as unsatisfactory effects, low classification accuracy, and weak adaptive ability. This method separates image feature extraction and classification into two steps for classification operation. The deep learning model has a powerful learning ability, which integrates the feature extraction and classification process into a whole to complete the image classification test, which can effectively improve the image classification accuracy. However, this method has the following problems in the application process: first, it is impossible to effectively approximate the complex functions in the deep learning model. Second, the deep learning model comes with a low classifier with low accuracy. So, this paper introduces the idea of sparse representation into the architecture of the deep learning network and comprehensively utilizes the sparse representation of well multidimensional data linear decomposition ability and the deep structural advantages of multilayer nonlinear mapping to complete the complex function approximation in the deep learning model. And a sparse representation classification method based on the optimized kernel function is proposed to replace the classifier in the deep learning model, thereby improving the image classification effect. Therefore, this paper proposes an image classification algorithm based on the stacked sparse coding depth learning model-optimized kernel function nonnegative sparse representation. The experimental results show that the proposed method not only has a higher average accuracy than other mainstream methods but also can be good adapted to various image databases. Compared with other deep learning methods, it can better solve the problems of complex function approximation and poor classifier effect, thus further improving image classification accuracy.

Homomorphic encryption (HE) is the ultimate tool for performing secure computations even in untrusted environments. Application of HE for deep learning (DL) inference is an active area of research, given the fact that DL models are often deployed in untrusted environments (e.g., third-party servers) yet inferring on private data. However, existing HE libraries [somewhat (SWHE), leveled (LHE) or fully homomorphic (FHE)] suffer from extensive computational and memory overhead. Few performance optimized high-speed homomorphic libraries are either suffering from certain approximation issues leading to decryption errors or proven to be insecure according to recent published attacks. In this article, we propose architectural tricks to achieve performance speedup for encrypted DL inference developed with exact HE schemes without any approximation or decryption error in homomorphic computations. The main idea is to apply quantization and suitable data packing in the form of bitslicing to reduce the costly noise handling operation, Bootstrapping while achieving a functionally correct and highly parallel DL pipeline with a moderate memory footprint. Experimental evaluation on the MNIST dataset shows a significant ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$37\times$ </tex-math></inline-formula> ) speedup over the nonbitsliced versions of the same architecture. Low memory bandwidths (700 MB) of our design pipelines further highlight their promise toward scaling over larger gamut of Edge-AI analytics use cases.

Remote-sensing images comprise massive amount of spatial and semantic data that can be employed for several applications. Presently, deep learning (DL) models for RS image processing become a familiar research area. Due to the advancements of recent satellite imaging sensors , the issue of huge amount of data processing becomes a challenging problem. To accomplish this, deep transfer learning (DTL) models are developed to resolve the semantic gap among various datasets This study develops a new DTL-based fusion model for environmental remote-sensing image classification, called DTLF-ERSIC technique. The proposed technique focuses on the design of fusion model to combine multiple feature vectors and thereby attains maximum classification performance. The DTLF-ERSIC technique incorporates the entropy-based fusion of three feature extraction techniques, namely, Discrete Local Binary Pattern (DLBP), Residual Network (ResNet50), and EfficientNet models. Besides, a rain optimization algorithm (ROA) with fuzzy rule-based classifier (FRC) is applied to predict the class labels of the test RS images and shows the novelty of the work. A comprehensive experimental analysis of the DTLF-ERSIC technique takes place on benchmark dataset and examined the results in terms of different performance measures. The simulation results reported the supremacy of the DTLF-ERSIC technique over the recent state-of-art techniques.

The past few years have seen a surge of applying Deep Learning (DL) models for a wide array of tasks such as image classification, object detection, machine translation, etc. While DL models provide an opportunity to solve otherwise intractable tasks, their adoption relies on them being optimized to meet latency and resource requirements. Benchmarking is a key step in this process but has been hampered in part due to the lack of representative and up-to-date benchmarking suites. This is exacerbated by the fast-evolving pace of DL models. This paper proposes DLBricks, a composable benchmark generation design that reduces the effort of developing, maintaining, and running DL benchmarks on CPUs. DLBricks decomposes DL models into a set of unique runnable networks and constructs the original model's performance using the performance of the generated benchmarks. DLBricks leverages two key observations: DL layers are the performance building blocks of DL models and layers are extensively repeated within and across DL models. Since benchmarks are generated automatically and the benchmarking time is minimized, DLBricks can keep up-to-date with the latest proposed models, relieving the pressure of selecting representative DL models. Moreover, DLBricks allows users to represent proprietary models within benchmark suites. We evaluate DLBricks using $50$ MXNet models spanning $5$ DL tasks on $4$ representative CPU systems. We show that DLBricks provides an accurate performance estimate for the DL models and reduces the benchmarking time across systems (e.g. within $95\%$ accuracy and up to $4.4\times$ benchmarking time speedup on Amazon EC2 c5.xlarge).

ReActHE: A homomorphic encryption friendly deep neural network for privacy-preserving biomedical prediction

HEFactory: A symbolic execution compiler for privacy-preserving Deep Learning with Homomorphic Encryption

This chapter discusses two popular homomorphic public key encryption (PKE) techniques, i.e., Paillier PKE and Boneh-Goh-Nissim (BGN) PKE, which serve as the preliminary for building most privacy-enhancing aggregation techniques in the rest chapters of this monograph. In addition, the Java source codes of the two homomorphic PKEs are also provided for the interested readers to better understand and implement them. In order to attain privacy-preserving data aggregation in smart grid communications, we need to first understand the homomorphic public key encryption (HPKE) techniques [1, 2, 3, 4, 5, 6, 7]. Different from the general public key encryption algorithms [8, 9], HPKE further holds an additional “homomorphic” property, which makes the privacy-preserving data aggregation possible. As shown in Fig. 2.1, when we directly operate over two encrypted data E(x) and E(y) with some operation “•”, we can gain E(x ∘ y) = E(x)•E(y). In most HPKE cases, the operations “•” and “∘” are respectively referred to the common multiplication “×” and addition “+” operations. Because most of privacy-enhancing aggregation techniques illustrated in this monograph are based on Paillier public key encryption [2] and Boneh-Goh-Nissim (BGN) public key encryption [6], in this chapter, we first take a close look at these two popular homomorphic encryption techniques. Note that, both of them are randomized encryption algorithms [1], i.e., in addition to the plaintext as the input of encryption, a random number is also input for achieving semantic security.

Decentralized image classification plays a key role in various scenarios due to its attractive properties, including tolerating high network latency and less prone to single-point failures. Unfortunately, training such a decentralized image classification model is more vulnerable to data privacy leaks compared to other distributed training frameworks. Existing efforts exclusively use differential privacy as the cornerstone to alleviate the threat to data privacy. However, differential privacy is implemented at the expense of accuracy, which goes against our motivation for designing an image classification model without loss of accuracy. To address this problem, we propose D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -MHE, the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">first</i> secure and efficient decentralized training framework with lossless precision. Inspired by the latest developments in the homomorphic encryption technology, we design a multiparty version of Brakerski-Fan-Vercauteren (BFV), one of the most advanced cryptosystems, and use it to implement private gradient updates of users' local models. D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -MHE can reduce the communication complexity of general Secure Multiparty Computation (MPC) tasks from quadratic to linear in the number of users, making it very suitable and scalable for large-scale decentralized learning systems. Moreover, D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -MHE provides strict semantic security protection even if the majority of users are dishonest with collusion. We conduct extensive experiments on MNIST, CIFAR-10, and ImageNet to demonstrate the superiority of D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -MHE. Experimental results show that D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -MHE achieves up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$5.5\times $ </tex-math></inline-formula> reduction in computation overhead, and at least <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$12\times $ </tex-math></inline-formula> reduction in communication overhead compared to existing schemes.

Statistical confidentiality focuses on protecting data to preserve its analytical value while preventing identity exposure, ensuring privacy and security in any system handling sensitive information. Homomorphic encryption allows computations on encrypted data without revealing it to anyone other than an owner or an authorized collector. When combined with other techniques, homomorphic encryption offers an ideal solution for ensuring statistical confidentiality. TFHE (Fast Fully Homomorphic Encryption over the Torus) is a fully homomorphic encryption scheme that supports efficient homomorphic operations on Booleans and integers. Building on TFHE, Zama’s Concrete project offers an open-source compiler that translates high-level Python code (version 3.9 or higher) into secure homomorphic computations. This study examines the feasibility of the Concrete compiler to perform core statistical analyses on encrypted data. We implement traditional algorithms for core statistical measures including the mean, variance, and five-point summary on encrypted datasets. Additionally, we develop a bitonic sort implementation to support the five-point summary. All implementations are executed within the Concrete framework, leveraging its built-in optimizations. Their performance is systematically evaluated by measuring circuit complexity, programmable bootstrapping count (PBS), compilation time, and execution time. We compare these results to findings from previous studies wherever possible. The results show that the complexity of sorting and statistical computations on encrypted data with the Concrete implementation of TFHE increases rapidly, and the size and range of data that can be accommodated is small for most applications. Nevertheless, this work reinforces the theoretical promise of Fully Homomorphic Encryption (FHE) for statistical analysis and highlights a clear path forward: the development of optimized, FHE-compatible algorithms.

PurposeTo conduct a head-to-head comparison between deep learning (DL) and radiomics models across institutions for predicting microvascular invasion (MVI) in hepatocellular carcinoma (HCC) and to investigate the model robustness and generalizability through rigorous internal and external validation.MethodsThis retrospective study included 2304 preoperative images of 576 HCC lesions from two centers, with MVI status determined by postoperative histopathology. We developed DL and radiomics models for predicting the presence of MVI using B-mode ultrasound, contrast-enhanced ultrasound (CEUS) at the arterial, portal, and delayed phases, and a combined modality (B + CEUS). For radiomics, we constructed models with enlarged vs. original regions of interest (ROIs). A cross-validation approach was performed by training models on one center’s dataset and validating the other, and vice versa. This allowed assessment of the validity of different ultrasound modalities and the cross-center robustness of the models. The optimal model combined with alpha-fetoprotein (AFP) was also validated. The head-to-head comparison was based on the area under the receiver operating characteristic curve (AUC).ResultsThirteen DL models and 25 radiomics models using different ultrasound modalities were constructed and compared. B + CEUS was the optimal modality for both DL and radiomics models. The DL model achieved AUCs of 0.802–0.818 internally and 0.667–0.688 externally across the two centers, whereas radiomics achieved AUCs of 0.749–0.869 internally and 0.646–0.697 externally. The radiomics models showed overall improvement with enlarged ROIs (P < 0.05 for both CEUS and B + CEUS modalities). The DL models showed good cross-institutional robustness (P > 0.05 for all modalities, 1.6–2.1% differences in AUC for the optimal modality), whereas the radiomics models had relatively limited robustness across the two centers (12% drop-off in AUC for the optimal modality). Adding AFP improved the DL models (P < 0.05 externally) and well maintained the robustness, but did not benefit the radiomics model (P > 0.05).ConclusionCross-institutional validation indicated that DL demonstrated better robustness than radiomics for preoperative MVI prediction in patients with HCC, representing a promising solution to non-standardized ultrasound examination procedures.

Deep learning (DL) has shown outstanding performances in many fields, including remote sensing (RS). DL is turning into an essential tool for the RS research community. Recently, many cloud platforms have been developed to provide access to large-scale computing capacity, consequently permitting the usage of DL architectures as a service. However, this opened the door to new challenges associated with the privacy and security of data. The RS data used to train the DL algorithms have several privacy requirements. Some of them need a high level of confidentiality, such as satellite images related to public security with high spatial resolutions. Moreover, satellite images are usually protected by copyright, and the owner may strictly refuse to share them. Therefore, privacy-preserving deep learning (PPDL) techniques are a possible solution to this problem. PPDL enables training DL on encrypted data without revealing the original plaintext. This study proposes a hybrid PPDL approach for object classification for very-high-resolution satellite images. The proposed encryption scheme combines Paillier homomorphic encryption (PHE) and somewhat homomorphic encryption (SHE). This combination aims to enhance the encryption of satellite images while ensuring a good runtime and high object classification accuracy. The method proposed to encrypt images is maintained through the public keys of PHE and SHE. Experiments were conducted on real-world high-resolution satellite images acquired using the SPOT6 and SPOT7 satellites. Four different CNN architectures were considered, namely ResNet50, InceptionV3, DenseNet169, and MobileNetV2. The results showed that the loss in classification accuracy after applying the proposed encryption algorithm ranges from 2% to 3.5%, with the best validation accuracy on the encrypted dataset reaching 92%.

The aim of this study was to distinguish canine lymphoma from other diseases,particularly reactive lymphoid hyperplasia (RLH), based on fine needle aspiration (FNA)images. We developed four deep learning models based on Vision Transformer (ViT) andInception-v3, which were pre-trained image classification models. The two models out offour were ViT and Inception-v3, and the remained were the two types of combination, i.e.,ensemble learning models, of ViT and Inception-v3; the mean of class probabilities of ViTand Inception-v3 (Ensemble model A; MEAN) and the maximum probabilities of ViT andInception-v3 (Ensemble model B; MAX). A total of 2,290 FNA images of canine lymphoma and871 FNA images of RLH were analyzed. The FNA images were obtained from the twenty-fiveslides of fourteen lymphoma cases and eight slides of seven RLH cases in two hospitals.Three types of training and test datasets were prepared from the above image datasets forfair evaluation of the models. Three deep learning-based image classification models(Inception-v3 and the two ensemble models) attained high performance of >80% accuracy,recall and area under the curve (AUC) values for all three datasets. ViT did not archivehigh performance, except the precision (>0.85). This study is an example of showingpotentials of deep learning models through image classification problem in caninelymphoma.

A robust and lightweight deep attention multiple instance learning algorithm for predicting genetic alterations.

Abstract: Encryption is essential in preventing unauthorized access to sensitive data in light of the growing concerns about data security in cloud computing. Homomorphic encryption promises to enable secure calculations on encrypted data without the need for decryption, particularly for cloud-based operations. To evaluate the effectiveness and applicability of several homomorphic encryption algorithms for safe cloud computing, we compare and contrast them in this research paper. Partially homomorphic encryption (PHE), somewhat homomorphic encryption (SHE), and fully homomorphic encryption (FHE) are the three basic homomorphic encryption subtypes that we examine. The implications of this study can aid cloud service providers and organizations in selecting the most appropriate homomorphic encryption scheme based on their specific security requirements and performance considerations. The research contributes to the ongoing efforts to enhance data privacy in cloud computing environments, opening new possibilities for secure data processing in an increasingly connected digital world. The exploration of homomorphic encryption schemes in this study opens new avenues for research and development in the field of cryptographic techniques. As technology continues to evolve, so too must our approaches to safeguarding data. This research serves as a catalyst for further innovations in homomorphic encryption algorithms, enabling even more efficient and robust methods for secure data processing in cloud environments and beyond. The insights derived from this research paper not only empower cloud service providers and organizations to make informed decisions about selecting the most appropriate homomorphic encryption scheme but also contribute to the broader mission of fortifying data privacy and security in cloud computing.