Bridging Knowledge Gap Between Image Inpainting and Large-Area Visible Watermark Removal
This work addresses the challenge of robust visible watermark removal for applications in digital media and security, representing an incremental advancement by adapting pre-trained inpainting models.
The paper tackles the problem of removing large-area visible watermarks from images, which existing deep neural network models struggle with due to dependency on high-quality watermark masks, and achieves significant performance improvements over state-of-the-art methods as demonstrated in experiments on synthesized and real-world datasets.
Visible watermark removal which involves watermark cleaning and background content restoration is pivotal to evaluate the resilience of watermarks. Existing deep neural network (DNN)-based models still struggle with large-area watermarks and are overly dependent on the quality of watermark mask prediction. To overcome these challenges, we introduce a novel feature adapting framework that leverages the representation modeling capacity of a pre-trained image inpainting model. Our approach bridges the knowledge gap between image inpainting and watermark removal by fusing information of the residual background content beneath watermarks into the inpainting backbone model. We establish a dual-branch system to capture and embed features from the residual background content, which are merged into intermediate features of the inpainting backbone model via gated feature fusion modules. Moreover, for relieving the dependence on high-quality watermark masks, we introduce a new training paradigm by utilizing coarse watermark masks to guide the inference process. This contributes to a visible image removal model which is insensitive to the quality of watermark mask during testing. Extensive experiments on both a large-scale synthesized dataset and a real-world dataset demonstrate that our approach significantly outperforms existing state-of-the-art methods. The source code is available in the supplementary materials.