Robust watermarking with double detector-discriminator approach

TL;DR

This paper introduces a deep learning-based watermarking framework with a novel double detector-discriminator scheme, achieving high robustness and transparency, especially against JPEG compression and various image distortions.

Contribution

It presents a new neural network approach for watermarking that outperforms existing methods in robustness, introduces a double detector-discriminator scheme, and offers a flexible method to balance image quality and robustness.

Findings

Bit accuracy of at least 0.86 across distortions

Achieved 0.90 bit accuracy for JPEG compression

Outperforms recent methods in robustness and transparency

Abstract

In this paper we present a novel deep framework for a watermarking - a technique of embedding a transparent message into an image in a way that allows retrieving the message from a (perturbed) copy, so that copyright infringement can be tracked. For this technique, it is essential to extract the information from the image even after imposing some digital processing operations on it. Our framework outperforms recent methods in the context of robustness against not only spectrum of attacks (e.g. rotation, resizing, Gaussian smoothing) but also against compression, especially JPEG. The bit accuracy of our method is at least 0.86 for all types of distortions. We also achieved 0.90 bit accuracy for JPEG while recent methods provided at most 0.83. Our method retains high transparency and capacity as well. Moreover, we present our double detector-discriminator approach - a scheme to detect and discriminate if the image contains the embedded message or not, which is crucial for real-life watermarking systems and up to now was not investigated using neural networks. With this, we design a testing formula to validate our extended approach and compared it with a common procedure. We also present an alternative method of balancing between image quality and robustness on attacks which is easily applicable to the framework.