RoWSFormer: A Robust Watermarking Framework with Swin Transformer for Enhanced Geometric Attack Resilience
This addresses robustness issues in watermarking for digital content protection, though it appears incremental as it adapts existing transformer architectures to a specific domain.
The paper tackles the problem of geometric attack resilience in digital image watermarking by proposing RoWSFormer, a framework based on Swin Transformer that captures global relationships. Experimental results show it improves PSNR by 3 dB for non-geometric attacks and over 6 dB for geometric attacks, with extraction accuracy exceeding 97%.
In recent years, digital watermarking techniques based on deep learning have been widely studied. To achieve both imperceptibility and robustness of image watermarks, most current methods employ convolutional neural networks to build robust watermarking frameworks. However, despite the success of CNN-based watermarking models, they struggle to achieve robustness against geometric attacks due to the limitations of convolutional neural networks in capturing global and long-range relationships. To address this limitation, we propose a robust watermarking framework based on the Swin Transformer, named RoWSFormer. Specifically, we design the Locally-Channel Enhanced Swin Transformer Block as the core of both the encoder and decoder. This block utilizes the self-attention mechanism to capture global and long-range information, thereby significantly improving adaptation to geometric distortions. Additionally, we construct the Frequency-Enhanced Transformer Block to extract frequency domain information, which further strengthens the robustness of the watermarking framework. Experimental results demonstrate that our RoWSFormer surpasses existing state-of-the-art watermarking methods. For most non-geometric attacks, RoWSFormer improves the PSNR by 3 dB while maintaining the same extraction accuracy. In the case of geometric attacks (such as rotation, scaling, and affine transformations), RoWSFormer achieves over a 6 dB improvement in PSNR, with extraction accuracy exceeding 97\%.