Advances in Digital Watermarking and Image Forensics

The field of digital watermarking and image forensics is rapidly evolving, with a focus on developing robust and secure methods for attributing and verifying the authenticity of digital content. Recent research has highlighted the limitations of existing watermarking schemes, particularly in the context of generative AI models. Innovations in this area include the use of diffusion models, chaos-based cryptographic techniques, and multi-modal watermarking approaches. These advancements have significant implications for the protection of intellectual property and the prevention of malicious content dissemination. Noteworthy papers in this area include:

  • Imperceptible but Forgeable, which demonstrates the vulnerability of current watermarking paradigms to forgery attacks.
  • TraceMark-LDM, which introduces a novel algorithm for attributing generated images while guaranteeing non-destructive performance.
  • RoSMM, which presents a robust and secure multi-modal watermarking framework for diffusion models.

Sources

Imperceptible but Forgeable: Practical Invisible Watermark Forgery via Diffusion Models

DF-Net: The Digital Forensics Network for Image Forgery Detection

DF2023: The Digital Forensics 2023 Dataset for Image Forgery Detection

A Novel Chaos-Based Cryptographic Scrambling Technique to Secure Medical Images

TraceMark-LDM: Authenticatable Watermarking for Latent Diffusion Models via Binary-Guided Rearrangement

Construction of Hyperchaotic Maps Based on 3D-CCC and its Applications in Image Encryption

MultiNeRF: Multiple Watermark Embedding for Neural Radiance Fields

RoSMM: A Robust and Secure Multi-Modal Watermarking Framework for Diffusion Models

Built with on top of