PrivacyBench: Privacy Isn't Free in Hybrid Privacy-Preserving Vision Systems

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Obiefuna, Nnaemeka, Oyeneye, Samuel, Odunaiya, Similoluwa, Oyelaja, Iremide, Kolawole, Steven
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917286471794688
author Obiefuna, Nnaemeka
Oyeneye, Samuel
Odunaiya, Similoluwa
Oyelaja, Iremide
Kolawole, Steven
author_facet Obiefuna, Nnaemeka
Oyeneye, Samuel
Odunaiya, Similoluwa
Oyelaja, Iremide
Kolawole, Steven
contents Privacy preserving machine learning deployments in sensitive deep learning applications; from medical imaging to autonomous systems; increasingly require combining multiple techniques. Yet, practitioners lack systematic guidance to assess the synergistic and non-additive interactions of these hybrid configurations, relying instead on isolated technique analysis that misses critical system level interactions. We introduce PrivacyBench, a benchmarking framework that reveals striking failures in privacy technique combinations with severe deployment implications. Through systematic evaluation across ResNet18 and ViT models on medical datasets, we uncover that FL + DP combinations exhibit severe convergence failure, with accuracy dropping from 98% to 13% while compute costs and energy consumption substantially increase. In contrast, FL + SMPC maintains near-baseline performance with modest overhead. Our framework provides the first systematic platform for evaluating privacy-utility-cost trade-offs through automated YAML configuration, resource monitoring, and reproducible experimental protocols. PrivacyBench enables practitioners to identify problematic technique interactions before deployment, moving privacy-preserving computer vision from ad-hoc evaluation toward principled systems design. These findings demonstrate that privacy techniques cannot be composed arbitrarily and provide critical guidance for robust deployment in resource-constrained environments.
format Preprint
id arxiv_https___arxiv_org_abs_2602_18900
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle PrivacyBench: Privacy Isn't Free in Hybrid Privacy-Preserving Vision Systems
Obiefuna, Nnaemeka
Oyeneye, Samuel
Odunaiya, Similoluwa
Oyelaja, Iremide
Kolawole, Steven
Cryptography and Security
Computer Vision and Pattern Recognition
I.4.9; C.2.4; K.6.5
Privacy preserving machine learning deployments in sensitive deep learning applications; from medical imaging to autonomous systems; increasingly require combining multiple techniques. Yet, practitioners lack systematic guidance to assess the synergistic and non-additive interactions of these hybrid configurations, relying instead on isolated technique analysis that misses critical system level interactions. We introduce PrivacyBench, a benchmarking framework that reveals striking failures in privacy technique combinations with severe deployment implications. Through systematic evaluation across ResNet18 and ViT models on medical datasets, we uncover that FL + DP combinations exhibit severe convergence failure, with accuracy dropping from 98% to 13% while compute costs and energy consumption substantially increase. In contrast, FL + SMPC maintains near-baseline performance with modest overhead. Our framework provides the first systematic platform for evaluating privacy-utility-cost trade-offs through automated YAML configuration, resource monitoring, and reproducible experimental protocols. PrivacyBench enables practitioners to identify problematic technique interactions before deployment, moving privacy-preserving computer vision from ad-hoc evaluation toward principled systems design. These findings demonstrate that privacy techniques cannot be composed arbitrarily and provide critical guidance for robust deployment in resource-constrained environments.
title PrivacyBench: Privacy Isn't Free in Hybrid Privacy-Preserving Vision Systems
topic Cryptography and Security
Computer Vision and Pattern Recognition
I.4.9; C.2.4; K.6.5
url https://arxiv.org/abs/2602.18900