Image Data Hiding in Neural Compressed Latent Representations

TL;DR

This paper introduces a neural network-based image data hiding method that embeds secrets into compressed latent representations, achieving high image quality, secrecy, and robustness with significantly faster embedding speeds.

Contribution

It presents a novel end-to-end framework combining neural compression and data hiding, improving secrecy, robustness, and speed over existing methods.

Findings

Achieves high image quality and bit accuracy

Offers superior secrecy and competitive robustness

Speeds up embedding by over 50 times

Abstract

We propose an end-to-end learned image data hiding framework that embeds and extracts secrets in the latent representations of a generic neural compressor. By leveraging a perceptual loss function in conjunction with our proposed message encoder and decoder, our approach simultaneously achieves high image quality and high bit accuracy. Compared to existing techniques, our framework offers superior image secrecy and competitive watermarking robustness in the compressed domain while accelerating the embedding speed by over 50 times. These results demonstrate the potential of combining data hiding techniques and neural compression and offer new insights into developing neural compression techniques and their applications.