Distortion Agnostic Deep Watermarking

By Xiyang Luo, Ruohan Zhan, Huiwen Chang, Feng Yang, Peyman Milanfar. 2020

217 citations recorded by the source corpus.

Watermarking is the process of embedding information into an image that can survive under distortions, while requiring the encoded image to have little or no perceptual difference with the original image. Recently, deep learning-based methods achieved impressive results in both visual quality and message payload under a wide variety of image distortions. However, these methods all require differentiable models for the image distortions at training time, and may generalize poorly to unknown distortions. This is undesirable since the types of distortions applied to watermarked images are usually unknown and non-differentiable. In this paper, we propose a new framework for distortion-agnostic watermarking, where the image distortion is not explicitly modeled during training. Instead, the robustness of our system comes from two sources: adversarial training and channel coding. Compared to training on a fixed set of distortions and noise levels, our method achieves comparable or better results on distortions available during training, and better performance overall on unknown distortions.

Subfields: Computer Vision and Pattern Recognition, Privacy & Security

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Privacy & Security

Citations in the corpus

Artificial Fingerprinting for Generative Models: Rooting Deepfake Attribution in Training Data (2020, methodology): “Recent works replace hand-crafted hiding procedures with neural network encoding [4, 19, 49, 62, 59, 45, 34].”
The Stable Signature: Rooting Watermarks in Latent Diffusion Models (2023, methodology): “More recently, deep learning encoder/extractor alternatives like HiDDeN [2, 44, 51, 100, 108] or iterative methods by Vukotić et al.”
HiNet: Deep Image Hiding by Invertible Network (2021, background): “[20] further enhanced the robustness of network to unknown distortions by replacing a fixed set of distortions by a generator.”
Generating Images with Multimodal Language Models (2023, background): “These harms may possibly be mitigated by introducing watermarking into generated images [37, 70], or by deploying systems to detect generated images [12].”