InvZW: Invariant Feature Learning via Noise-Adversarial Training for Robust Image Zero-Watermarking

TL;DR

This paper presents InvZW, a deep learning-based zero-watermarking method that learns distortion-invariant features for robust and unaltered image watermarking, outperforming existing techniques in robustness and generalization.

Contribution

The paper introduces a novel noise-adversarial training framework for invariant feature learning in zero-watermarking, enhancing robustness without altering the original image.

Findings

Achieves state-of-the-art robustness in watermark recovery across diverse distortions.

Outperforms existing self-supervised and deep watermarking methods in generalization.

Demonstrates superior feature stability and watermark accuracy in extensive experiments.

Abstract

This paper introduces a novel deep learning framework for robust image zero-watermarking based on distortion-invariant feature learning. As a zero-watermarking scheme, our method leaves the original image unaltered and learns a reference signature through optimization in the feature space. The proposed framework consists of two key modules. In the first module, a feature extractor is trained via noise-adversarial learning to generate representations that are both invariant to distortions and semantically expressive. This is achieved by combining adversarial supervision against a distortion discriminator and a reconstruction constraint to retain image content. In the second module, we design a learning-based multibit zero-watermarking scheme where the trained invariant features are projected onto a set of trainable reference codes optimized to match a target binary message. Extensive experiments on diverse image datasets and a wide range of distortions show that our method achieves state-of-the-art robustness in both feature stability and watermark recovery. Comparative evaluations against existing self-supervised and deep watermarking techniques further highlight the superiority of our framework in generalization and robustness.