RL-MoE: An Image-Based Privacy Preserving Approach In Intelligent Transportation System

TL;DR

RL-MoE introduces a privacy-preserving framework converting sensitive images into descriptive text using a Mixture-of-Experts and Reinforcement Learning, enhancing privacy and data utility in intelligent transportation systems.

Contribution

The paper presents RL-MoE, a novel approach combining MoE architecture and RL to generate privacy-preserving textual descriptions from images, improving privacy protection and semantic richness.

Findings

Reduces replay attack success rate to 9.4% on CFP-FP dataset

Generates richer textual descriptions than baseline methods

Provides a scalable solution for privacy in ITS applications

Abstract

The proliferation of AI-powered cameras in Intelligent Transportation Systems (ITS) creates a severe conflict between the need for rich visual data and the right to privacy. Existing privacy-preserving methods, such as blurring or encryption, are often insufficient due to creating an undesirable trade-off where either privacy is compromised against advanced reconstruction attacks or data utility is critically degraded. To resolve this challenge, we propose RL-MoE, a novel framework that transforms sensitive visual data into privacy-preserving textual descriptions, eliminating the need for direct image transmission. RL-MoE uniquely combines a Mixture-of-Experts (MoE) architecture for nuanced, multi-aspect scene decomposition with a Reinforcement Learning (RL) agent that optimizes the generated text for a dual objective of semantic accuracy and privacy preservation. Extensive experiments demonstrate that RL-MoE provides superior privacy protection, reducing the success rate of replay attacks to just 9.4\% on the CFP-FP dataset, while simultaneously generating richer textual content than baseline methods. Our work provides a practical and scalable solution for building trustworthy AI systems in privacy-sensitive domains, paving the way for more secure smart city and autonomous vehicle networks.