SenGuard: A Novel Processing In-Sensor Method for Privacy-Enhanced Smart Imaging

Neeraj Solanki, Sepehr Tabrizchi, Ali Shafiee Sarvestani, Shaahin Angizi, Arman Roohi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper introduces SenGuard, a novel processing-in-sensor scheme to enhance imager privacy by integrating sensing and neural network computations directly within the pixel array. We demonstrate SenGuard's effectiveness against NeuralRecon, an attack combining side-channel techniques with generative adversarial networks to reconstruct input images. Evaluations across multiple datasets show SenGuard reduces attacker accuracy to as low as while maintaining classification performance with only a accuracy reduction compared to baseline models. SenGuard achieves 3.28 TOps/W energy efficiency, while making direct reconstruction of captured data extremely difficult.

Original languageEnglish (US)
Title of host publicationGLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025
PublisherAssociation for Computing Machinery
Pages587-592
Number of pages6
ISBN (Electronic)9798400714962
DOIs
StatePublished - Jun 29 2025
Event35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025 - New Orleans, United States
Duration: Jun 30 2025Jul 2 2025

Publication series

NameProceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI

Conference

Conference35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025
Country/TerritoryUnited States
CityNew Orleans
Period6/30/257/2/25

All Science Journal Classification (ASJC) codes

  • General Engineering

Fingerprint

Dive into the research topics of 'SenGuard: A Novel Processing In-Sensor Method for Privacy-Enhanced Smart Imaging'. Together they form a unique fingerprint.

Cite this