Visible Watermark Removal via Self-calibrated Localization and Background Refinement
This work addresses the incremental improvement of watermark removal techniques for copyright protection in images.
The paper tackles the problem of incomplete watermark detection and degraded texture quality in visible watermark removal by proposing a two-stage multi-task network with self-calibrated localization and background refinement, achieving effectiveness demonstrated through extensive experiments on two datasets.
Superimposing visible watermarks on images provides a powerful weapon to cope with the copyright issue. Watermark removal techniques, which can strengthen the robustness of visible watermarks in an adversarial way, have attracted increasing research interest. Modern watermark removal methods perform watermark localization and background restoration simultaneously, which could be viewed as a multi-task learning problem. However, existing approaches suffer from incomplete detected watermark and degraded texture quality of restored background. Therefore, we design a two-stage multi-task network to address the above issues. The coarse stage consists of a watermark branch and a background branch, in which the watermark branch self-calibrates the roughly estimated mask and passes the calibrated mask to background branch to reconstruct the watermarked area. In the refinement stage, we integrate multi-level features to improve the texture quality of watermarked area. Extensive experiments on two datasets demonstrate the effectiveness of our proposed method.