Perceptive self-supervised learning network for noisy image watermark removal
Chunwei Tian, Menghua Zheng, Bo Li, Yanning Zhang, Shichao Zhang, David Zhang
TL;DR
PSLNet addresses watermark removal from noisy images without requiring reference clean images by combining a self-supervised pairing strategy with a parallel two-branch architecture that separately and jointly tackle noise and watermark removal. The upper branch applies sequential denoising and watermark removal, while the lower branch follows a degradation-model approach to remove both factors simultaneously; interactions and a final fusion enhance structural integrity and pixel quality. A perceptual network based on a 16-layer VGG extracts texture cues to support a mixed loss that couples structural and perceptual objectives. Across diverse watermark types, transparencies, and noise levels, PSLNet outperforms several CNN baselines in PSNR, SSIM, and LPIPS, while maintaining reasonable complexity, demonstrating practical effectiveness for real-world watermark removal.
Abstract
Popular methods usually use a degradation model in a supervised way to learn a watermark removal model. However, it is true that reference images are difficult to obtain in the real world, as well as collected images by cameras suffer from noise. To overcome these drawbacks, we propose a perceptive self-supervised learning network for noisy image watermark removal (PSLNet) in this paper. PSLNet depends on a parallel network to remove noise and watermarks. The upper network uses task decomposition ideas to remove noise and watermarks in sequence. The lower network utilizes the degradation model idea to simultaneously remove noise and watermarks. Specifically, mentioned paired watermark images are obtained in a self supervised way, and paired noisy images (i.e., noisy and reference images) are obtained in a supervised way. To enhance the clarity of obtained images, interacting two sub-networks and fusing obtained clean images are used to improve the effects of image watermark removal in terms of structural information and pixel enhancement. Taking into texture information account, a mixed loss uses obtained images and features to achieve a robust model of noisy image watermark removal. Comprehensive experiments show that our proposed method is very effective in comparison with popular convolutional neural networks (CNNs) for noisy image watermark removal. Codes can be obtained at https://github.com/hellloxiaotian/PSLNet.