CLOAK: Contrastive Guidance for Latent Diffusion-Based Data Obfuscation

2025-12-12

AI Summary

Sensor time-series data in semi-honest settings is vulnerable to attribute inference attacks, yet existing obfuscation methods struggle to simultaneously ensure privacy preservation, downstream utility, and device adaptability. Method: We propose the first obfuscation framework integrating contrastive learning with latent diffusion models. It uniquely embeds contrastive learning into the latent diffusion process to decouple and controllably balance privacy protection and utility; supports multi-user differential privacy requirements without retraining; and introduces a lightweight adversarial guidance mechanism to significantly reduce computational overhead. Results: Extensive experiments on four time-series datasets and face images demonstrate that our method achieves superior trade-offs between privacy (robustness against attribute inference) and utility (classification/prediction accuracy), outperforming state-of-the-art approaches. Moreover, it enables efficient deployment on resource-constrained mobile IoT devices.

Technology Category

• Machine Learning: Privacy
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Data obfuscation is a promising technique for mitigating attribute inference attacks by semi-trusted parties with access to time-series data emitted by sensors. Recent advances leverage conditional generative models together with adversarial training or mutual information-based regularization to balance data privacy and utility. However, these methods often require modifying the downstream task, struggle to achieve a satisfactory privacy-utility trade-off, or are computationally intensive, making them impractical for deployment on resource-constrained mobile IoT devices. We propose Cloak, a novel data obfuscation framework based on latent diffusion models. In contrast to prior work, we employ contrastive learning to extract disentangled representations, which guide the latent diffusion process to retain useful information while concealing private information. This approach enables users with diverse privacy needs to navigate the privacy-utility trade-off with minimal retraining. Extensive experiments on four public time-series datasets, spanning multiple sensing modalities, and a dataset of facial images demonstrate that Cloak consistently outperforms state-of-the-art obfuscation techniques and is well-suited for deployment in resource-constrained settings.

Problem

• Addresses data privacy in IoT by obfuscating time-series sensor data.
• Improves privacy-utility trade-off without modifying downstream tasks.
• Enables efficient deployment on resource-constrained mobile devices.

Innovation

• Contrastive learning extracts disentangled representations for guidance
• Latent diffusion process retains utility while concealing private information
• Minimal retraining needed for diverse privacy-utility trade-offs