Robust Image Watermarking based on Cross-Attention and Invariant Domain Learning

TL;DR

This paper introduces a novel deep learning-based image watermarking method that leverages cross-attention and invariant domain learning to improve robustness and semantic embedding capabilities.

Contribution

It proposes a multi-head cross attention mechanism for embedding and an invariant domain learning approach to enhance watermark robustness and semantic relevance.

Findings

Improved robustness against image distortions.

Semantic-aware watermark embedding.

Enhanced generalization in watermark extraction.

Abstract

Image watermarking involves embedding and extracting watermarks within a cover image, with deep learning approaches emerging to bolster generalization and robustness. Predominantly, current methods employ convolution and concatenation for watermark embedding, while also integrating conceivable augmentation in the training process. This paper explores a robust image watermarking methodology by harnessing cross-attention and invariant domain learning, marking two novel, significant advancements. First, we design a watermark embedding technique utilizing a multi-head cross attention mechanism, enabling information exchange between the cover image and watermark to identify semantically suitable embedding locations. Second, we advocate for learning an invariant domain representation that encapsulates both semantic and noise-invariant information concerning the watermark, shedding light on promising avenues for enhancing image watermarking techniques.