Privacy-Shielded Image Compression: Defending Against Exploitation from Vision-Language Pretrained Models

TL;DR

This paper introduces Privacy-Shielded Image Compression (PSIC), a flexible method that protects image privacy from vision-language models during compression by producing multi-decoding options that hinder exploitation while maintaining image quality.

Contribution

The paper proposes a novel PSIC scheme with a CLTG module and UAEO optimization, enabling privacy protection at the compression stage without sacrificing perceptual quality or original functionality.

Findings

Effective in preventing VLP model exploitation

Maintains high perceptual image quality

Seamlessly integrates with existing LIC models

Abstract

The improved semantic understanding of vision-language pretrained (VLP) models has made it increasingly difficult to protect publicly posted images from being exploited by search engines and other similar tools. In this context, this paper seeks to protect users' privacy by implementing defenses at the image compression stage to prevent exploitation. Specifically, we propose a flexible coding method, termed Privacy-Shielded Image Compression (PSIC), that can produce bitstreams with multiple decoding options. By default, the bitstream is decoded to preserve satisfactory perceptual quality while preventing interpretation by VLP models. Our method also retains the original image compression functionality. With a customizable input condition, the proposed scheme can reconstruct the image that preserves its full semantic information. A Conditional Latent Trigger Generation (CLTG) module is proposed to produce bias information based on customizable conditions to guide the decoding process into different reconstructed versions, and an Uncertainty-Aware Encryption-Oriented (UAEO) optimization function is designed to leverage the soft labels inferred from the target VLP model's uncertainty on the training data. This paper further incorporates an adaptive multi-objective optimization strategy to obtain improved encrypting performance and perceptual quality simultaneously within a unified training process. The proposed scheme is plug-and-play and can be seamlessly integrated into most existing Learned Image Compression (LIC) models. Extensive experiments across multiple downstream tasks have demonstrated the effectiveness of our design.