EUVIMEDEUVIMED entdecken
Crossref

Encrypted Image Classification with Low Memory Footprint Using Fully Homomorphic Encryption

Lorenzo Rovida iD, Alberto Leporati iD

International Journal of Neural Systems · 2024-05 · 34 (05)

DOI: 10.1142/s0129065724500254 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Neural Systems, online erschienen am 22.03.2024. Verfasst von Lorenzo Rovida und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Informatics, Systems and Communication, University of Milan-Bicocca, Viale Sarca, 336, Milan, 20126, Italy. Hauptthemen laut MEDLINE: Computer Security, Machine Learning. Die Arbeit wurde bisher 34-mal zitiert (Crossref) und nennt 32 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Classifying images has become a straightforward and accessible task, thanks to the advent of Deep Neural Networks. Nevertheless, not much attention is given to the privacy concerns associated with sensitive data contained in images. In this study, we propose a solution to this issue by exploring an intersection between Machine Learning and cryptography. In particular, Fully Homomorphic Encryption (FHE) emerges as a promising solution, as it enables computations to be performed on encrypted data. We therefore propose a Residual Network implementation based on FHE which allows the classification of encrypted images, ensuring that only the user can see the result. We suggest a circuit which reduces the memory requirements by more than [Formula: see text] compared to the most recent works, while maintaining a high level of accuracy and a short computational time. We implement the circuit using the well-known Cheon–Kim–Kim–Song (CKKS) scheme, which enables approximate encrypted computations. We evaluate the results from three perspectives: memory requirements, computational time and calculations precision. We demonstrate that it is possible to evaluate an encrypted ResNet20 in less than five minutes on a laptop using approximately 15[Formula: see text]GB of memory, achieving an accuracy of 91.67% on the CIFAR-10 dataset, which is almost equivalent to the accuracy of the plain model (92.60%).

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Computer SecurityMachine LearningNeural Networks, Computer

Stichwörter der Autor:innen

Encrypted neural networkhomomorphic encryptionsecure machine learning

Autor:innen und Institutionen

  • Department of Informatics, Systems and Communication, University of Milan-Bicocca, Viale Sarca, 336, Milan, 20126, Italy.

ORCID-Profile: Lorenzo Rovida, Alberto Leporati

Zitieren

APA

Rovida, L., & Leporati, A. (2024). Encrypted Image Classification with Low Memory Footprint Using Fully Homomorphic Encryption. International Journal of Neural Systems, 34(05). https://doi.org/10.1142/s0129065724500254

Vancouver

Rovida L, Leporati A. Encrypted Image Classification with Low Memory Footprint Using Fully Homomorphic Encryption. International Journal of Neural Systems. 2024;34(05). doi:10.1142/s0129065724500254

BibTeX
@article{Rovida2024,
  author = {Rovida, Lorenzo and Leporati, Alberto},
  title = {Encrypted Image Classification with Low Memory Footprint Using Fully Homomorphic Encryption},
  journal = {International Journal of Neural Systems},
  year = {2024},
  volume = {34},
  number = {05},
  doi = {10.1142/s0129065724500254}
}
Kostenlos starten →