A Survey on Deep Learning Techniques for Privacy-Preserving

Tanuwidjaja, Harry Chandra; Choi, Rakyong; Kim, Kwangjo

doi:10.1007/978-3-030-30619-9_4

Cited by 35 publications

(24 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Akhtar et al [22] reviewed the adversarial example attacks and defenses on DL in the field of computer vision. Tanuwidjaja et al [23] and Boulemtafes et al [24] studied several privacypreserving techniques on DL. Yuan et al [25] presented a review on adversarial examples for DL, in which they summarize the adversarial example generation methods and discuss the corresponding defense methods.…”

Section: Introductionmentioning

confidence: 99%

Privacy and Security Issues in Deep Learning: A Survey

et al. 2021

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Privacy and Security Issues in Deep Learning: A Survey

et al. 2021

View full text Add to dashboard Cite

“…• To the best to our knowledge, this is the first PPDL survey paper to provides a complete discussion and analysis about PPDL on MLaaS. This paper is an extension of our previous version [1] by improving the classification of PPDL into a wider scope, adding latest publications of 28 papers in total with our analysis of the weakness of each method. We present comparison tables (Tables 8 and 9), based on the privacy parameter and performance level, respectively.…”

Section: Introductionmentioning

confidence: 94%

Privacy-Preserving Deep Learning on Machine Learning as a Service—a Comprehensive Survey

et al. 2020

Self Cite

View full text Add to dashboard Cite

The exponential growth of big data and deep learning has increased the data exchange traffic in society. Machine Learning as a Service, (MLaaS) which leverages deep learning techniques for predictive analytics to enhance decision-making, has become a hot commodity. However, the adoption of MLaaS introduces data privacy challenges for data owners and security challenges for deep learning model owners. Data owners are concerned about the safety and privacy of their data on MLaaS platforms, while MLaaS platform owners worry that their models could be stolen by adversaries who pose as clients. Consequently, Privacy-Preserving Deep Learning (PPDL) arises as a possible solution to this problem. Recently, several papers about PPDL for MLaaS have been published. However, to the best of our knowledge, no previous paper has summarized the existing literature on PPDL and its specific applicability to the MLaaS environment. In this paper, we present a comprehensive survey of privacypreserving techniques, starting from classical privacy-preserving techniques to well-known deep learning techniques. Additionally, we present a detailed description of PPDL and address the issue of using PPDL for MLaaS. Furthermore, we undertake detailed comparisons between state-of-the-art PPDL methods. Subsequently, we classify an adversarial model on PPDL by highlighting possible PPDL attacks and their potential solutions. Ultimately, our paper serves as a single point of reference for detailed knowledge on PPDL and its applicability to MLaaS environments for both new and experienced researchers.

show abstract

“…Due to the availability of huge amount of labeled data, and ability to work in a decentralized fashion, these techniques can be utilized for users' privacy preservation with enhanced usefulness. The heterogeneous federated transfer learning (HFTL) framework [298], privacy-preserving deep learning (PPDL) technique [299], deep transfer learning (DTL) method [300], adaptive privacy preserving federated learning (APPFL) method [301], block-chain-enable privacy preserving (BCEPP) architectures [302], [303], secure collaborative few-shot learning (SCFSL) framework [304], searchable encryption (SE) methods leveraging ciphertext-policy attribute-based encryption (CP-ABE) [305], [306], data resource protection solution leveraging smart contracts [307], improving cyber security solutions utilizing AI's potential [308], and computational intelligence based methods for information security [309], to name a few have already been used in practical applications related to the PPDP. Hence, devising robust and lightweight techniques which involve less parameters and can co-work with the traditional anonymization approaches to scale up privacy preservation with enhanced data utility is a promising area of research for the future.…”

Section: Promising Open Research Directionsmentioning

confidence: 99%