Abstract
Image watermarking is usually decomposed into three steps: (i) a feature vector is extracted from an image; (ii) it is modified to embed the watermark; (iii) and it is projected back into the image space while avoiding the creation of visual artefacts. This feature extraction is usually based on a classical image representation given by the Discrete Wavelet Transform or the Discrete Cosine Transform for instance. These transformations require very accurate synchronisation between the embedding and the detection and usually rely on various registration mechanisms for that purpose. This paper investigates a new family of transformation based on Deep Neural Networks trained with supervision for a classification task. Motivations come from the Computer Vision literature, which has demonstrated the robustness of these features against light geometric distortions. Also, adversarial sample literature provides means to implement the inverse transform needed in the third step above mentioned. As far as zero-bit watermarking is concerned, this paper shows that this approach is feasible as it yields a good quality of the watermarked images and an intrinsic robustness. We also tests more advanced tools from Computer Vision such as aggregation schemes with weak geometry and retraining with a dataset augmented with classical image processing attacks.
Highlights
Deep Learning (DL) has completely revolutionized the field of Computer Vision
As far as zero-bit watermarking is concerned, this paper shows that this approach is feasible as it yields a good quality of the watermarked images and an intrinsic robustness
It started with image classification [1] where a DL network is trained to recognize classes of images and it is spreading to any task of Computer Vision—object recognition [2], instance search [3], localization [4], similar image retrieval [5,6]
Summary
Deep Learning (DL) has completely revolutionized the field of Computer Vision. It started with image classification [1] where a DL network is trained to recognize classes of images and it is spreading to any task of Computer Vision—object recognition [2], instance search [3], localization [4], similar image retrieval [5,6]. Many Computer Vision works take a trained classification network off the shelf, keep the first layers as is and only re-train the deepest layers for their specific task [5] Following this trend, state-of-the-art image search algorithms tap the output of an internal layer and use them to derive a global descriptor of the image [5,6]. State-of-the-art image search algorithms tap the output of an internal layer and use them to derive a global descriptor of the image [5,6] This way, finding similar images boils down to finding close vectors in a Euclidean high dimensional space, where efficient fast search engines exist. Entropy 2020, 22, 198 time robust to valuemetric (e.g., JPEG, noise, filtering) and light geometric (e.g., cropping, rotation, scaling) distortions [7]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
