Supervised GAN Watermarking for Intellectual Property Protection

TL;DR

This paper introduces a novel supervised watermarking technique for GANs that embeds invisible, robust signatures into generated images for intellectual property protection, compatible with various architectures and resilient to post-processing.

Contribution

It presents a general, efficient method to embed invisible watermarks into GAN outputs using a pre-trained CNN decoder and modified training loss, enhancing IP protection.

Findings

Effective embedding of invisible watermarks in GAN-generated images.

Robustness against JPEG compression, noise, blurring, and color changes.

Compatibility with different GAN architectures and resolutions.

Abstract

We propose a watermarking method for protecting the Intellectual Property (IP) of Generative Adversarial Networks (GANs). The aim is to watermark the GAN model so that any image generated by the GAN contains an invisible watermark (signature), whose presence inside the image can be checked at a later stage for ownership verification. To achieve this goal, a pre-trained CNN watermarking decoding block is inserted at the output of the generator. The generator loss is then modified by including a watermark loss term, to ensure that the prescribed watermark can be extracted from the generated images. The watermark is embedded via fine-tuning, with reduced time complexity. Results show that our method can effectively embed an invisible watermark inside the generated images. Moreover, our method is a general one and can work with different GAN architectures, different tasks, and different resolutions of the output image. We also demonstrate the good robustness performance of the embedded watermark against several post-processing, among them, JPEG compression, noise addition, blurring, and color transformations.