analiza aiqbench
This study demonstrates that combining federated learning, client-side differential privacy, and INT8 quantization enables high-fidelity ECG anomaly detection on edge hardware without compromising privacy or performance. The research empirically validates that privacy mechanisms and model compression penalties are independent, allowing practitioners to deploy compact models on devices like the Raspberry Pi 4 with up to 44% latency reduction. Key takeaways include a recommended differential privacy budget of $\varepsilon=4$ and a unified architecture that satisfies legal-grade requirements while maintaining centralized baseline accuracy.
Privacy-Preserving Federated Autoencoder for ECG Anomaly Detection on Edge Devices
arXiv:2606.11556v1 Announce Type: cross Abstract: Continuous electrocardiography (ECG) monitoring could surface rhythm abnormalities before they escalate into cardiovascular events. However, a deployable system must satisfy three requirements simultaneously: legal-grade privacy (GDPR, HIPAA), real-time inference on constrained edge hardware, and detection quality under non-IID cross-hospital data. We design and evaluate an end-to-end federated system addressing all three for unsupervised 12-lead ECG anomaly detection on PTB-XL dataset, combining three autoencoder families (VanillaAE, ConvAE, VAE), Flower-based federated averaging (FedAvg) across ten simulated hospitals, client-side differentially private SGD (DP-SGD) with a R\'enyi-DP accountant, and 8-bit integer (INT8) post-training quantization with Raspberry Pi 4 benchmarking. Our main contributions are: an empirical characterization of how these mechanisms compose, practical DP-specific recommendations, and technical and security insights for a clinically sensitive setting. Federated learning matches or exceeds the centralized baseline across all architectures (ConvAE federated area under the ROC curve, AUROC, $0.782$), and an $\varepsilon$ sweep identifies $\varepsilon=4$ as the recommended clinical operating point. INT8 quantization roughly halves model size and cuts Pi 4 latency by up to $44%$ with $<0.12%$ AUROC loss. Crucially, DP and quantization penalties are empirically independent, so practitioners need not trade a strong privacy guarantee for a compact edge footprint. To our knowledge, this is the first system combining federated learning, formal $(\varepsilon,\delta)$-DP, unsupervised reconstruction-based detection, and quantized AArch64 deployment.
Abstract:Continuous electrocardiography (ECG) monitoring could surface rhythm abnormalities before they escalate into cardiovascular events. However, a deployable system must satisfy three requirements simultaneously: legal-grade privacy (GDPR, HIPAA), real-time inference on constrained edge hardware, and detection quality under non-IID cross-hospital data. We design and evaluate an end-to-end federated system addressing all three for unsupervised 12-lead ECG anomaly detection on PTB-XL dataset, combining three autoencoder families (VanillaAE, ConvAE, VAE), Flower-based federated averaging (FedAvg) across ten simulated hospitals, client-side differentially private SGD (DP-SGD) with a Rényi-DP accountant, and 8-bit integer (INT8) post-training quantization with Raspberry Pi 4 benchmarking. Our main contributions are: an empirical characterization of how these mechanisms compose, practical DP-specific recommendations, and technical and security insights for a clinically sensitive setting. Federated learning matches or exceeds the centralized baseline across all architectures (ConvAE federated area under the ROC curve, AUROC, $0.782$), and an $\varepsilon$ sweep identifies $\varepsilon=4$ as the recommended clinical operating point. INT8 quantization roughly halves model size and cuts Pi 4 latency by up to $44%$ with $<0.12%$ AUROC loss. Crucially, DP and quantization penalties are empirically independent, so practitioners need not trade a strong privacy guarantee for a compact edge footprint. To our knowledge, this is the first system combining federated learning, formal $(\varepsilon,\delta)$-DP, unsupervised reconstruction-based detection, and quantized AArch64 deployment.
9 pages, 4 figures, 6 tables. Preprint prepared in IEEE conference format. Submitted to: FLTA 2026
Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
I.2.6; I.5.4; J.3; C.2.4; C.3
arXiv:2606.11556 [cs.CR]
(or arXiv:2606.11556v1 [cs.CR] for this version)
https://doi.org/10.48550/arXiv.2606.11556 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 259 cuvinte
Subiecte conexe
[AINews] not much happened today
Latent Space · 3 oct. 2026
AI Swiftly Detects and Neutralises 6G Network Threats in Milliseconds
Quantum Zeitgeist · 10 iun. 2026
Reliable Federated TinyML Deployment for IoT Security
arXiv · 24 sept. 2026
Model Multiplicity for Adversarial Detection in Small Language Model Training on Edge Devices
arXiv · 9 iun. 2026