Hiding Images into Images with Real-world Robustness
This work addresses the problem of robust image watermarking for copyright protection in real-world scenarios, representing an incremental improvement with specific gains in attack resistance.
The paper tackles the vulnerability of existing image embedding networks to malicious attacks like JPEG compression and noise, making them unsuitable for real-world copyright protection, and introduces a generative deep network method that ensures high-quality extraction from attacked images, achieving superior performance against typical digital attacks and being the first to robustly hide three secret images.
The existing image embedding networks are basically vulnerable to malicious attacks such as JPEG compression and noise adding, not applicable for real-world copyright protection tasks. To solve this problem, we introduce a generative deep network based method for hiding images into images while assuring high-quality extraction from the destructive synthesized images. An embedding network is sequentially concatenated with an attack layer, a decoupling network and an image extraction network. The addition of decoupling network learns to extract the embedded watermark from the attacked image. We also pinpoint the weaknesses of the adversarial training for robustness in previous works and build our improved real-world attack simulator. Experimental results demonstrate the superiority of the proposed method against typical digital attacks by a large margin, as well as the performance boost of the recovered images with the aid of progressive recovery strategy. Besides, we are the first to robustly hide three secret images.