RWN: Robust Watermarking Network for Image Cropping Localization
This addresses the issue of image manipulation on social networks for users needing content verification, though it is incremental by extending watermarking to localization.
The paper tackles the problem of detecting and localizing malicious image cropping by introducing a robust watermarking network (RWN) that embeds and extracts watermarks to locate cropped areas, achieving high-accuracy crop localization and competitive tamper detection.
Image cropping can be maliciously used to manipulate the layout of an image and alter the underlying meaning. Previous image crop detection schemes only predicts whether an image has been cropped, ignoring which part of the image is cropped. This paper presents a novel robust watermarking network (RWN) for image crop localization. We train an anti-crop processor (ACP) that embeds a watermark into a target image. The visually indistinguishable protected image is then posted on the social network instead of the original image. At the recipient's side, ACP extracts the watermark from the attacked image, and we conduct feature matching on the original and extracted watermark to locate the position of the crop in the original image plane. We further extend our scheme to detect tampering attack on the attacked image. Besides, we explore a simple yet efficient method (JPEG-Mixup) to improve the generalization of JPEG robustness. According to our comprehensive experiments, RWN is the first to provide high-accuracy and robust image crop localization. Besides, the accuracy of tamper detection is comparable with many state-of-the-art passive-based methods.