Edge-Native Generative De-identification: Inversion-Free Flow for Privacy-Preserving Federated Skin Image Analysis

2026-01-31

AI Summary

This work addresses the challenge of simultaneously preserving patient identity privacy and maintaining pathological fidelity in federated learning for clinical dermatology. The authors propose a client-side, inversion-free generative de-identification framework based on Rectified Flow Transformers, enabling near-real-time, high-fidelity identity transformation. A novel “synthesis-guided segmentation” mechanism generates paired healthy–pathological synthetic twin images to extract erythema difference masks decoupled from biometric features. This approach safeguards privacy and prevents gradient leakage while retaining diagnostically critical information. Experiments on high-resolution clinical images demonstrate that cross-synthetic-identity segmentation achieves an IoU stability exceeding 0.67, with per-instance identity transformation completed in under 20 seconds, supporting practical deployment on edge devices.

Technology Category

• Machine Learning: Privacy
• Computer Vision: Bias, Fairness & Privacy
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

The deployment of Federated Learning (FL) for clinical dermatology is hindered by the competing requirements of protecting patient privacy and preserving diagnostic features. Traditional de-identification methods often degrade pathological fidelity, while standard generative editing techniques rely on computationally intensive inversion processes unsuitable for resource-constrained edge devices. We propose a framework for identity-agnostic pathology preservation that serves as a client-side privacy-preserving utility. By leveraging inversion-free Rectified Flow Transformers (FlowEdit), the system performs high-fidelity identity transformation in near real-time (less than 20s), facilitating local deployment on clinical nodes. We introduce a"Segment-by-Synthesis"mechanism that generates counterfactual healthy and pathological twin pairs locally. This enables the extraction of differential erythema masks that are decoupled from biometric markers and semantic artifacts (e.g. jewelry). Pilot validation on high-resolution clinical samples demonstrates an Intersection over Union (IoU) stability greater than 0.67 across synthetic identities. By generating privacy-compliant synthetic surrogates at the edge, this framework mitigates the risk of gradient leakage at the source, providing a secure pathway for high-precision skin image analysis in federated environments.

Problem

• privacy-preserving
• federated learning
• de-identification
• edge computing
• skin image analysis

Innovation

• Inversion-Free Generation
• Federated Learning
• Edge Computing
• Privacy-Preserving AI
• Rectified Flow Transformers