DeepForgeSeal: Latent Space-Driven Semi-Fragile Watermarking for Deepfake Detection Using Multi-Agent Adversarial Reinforcement Learning

TL;DR

DeepForgeSeal introduces a novel latent space-based semi-fragile watermarking framework utilizing multi-agent adversarial reinforcement learning to improve deepfake detection robustness and adaptability.

Contribution

It proposes a learnable watermark embedder in latent space combined with MAARL to balance robustness and fragility against manipulations.

Findings

Outperforms state-of-the-art methods on CelebA and CelebA-HQ benchmarks.

Achieves over 4.5% and 5.3% improvements under challenging scenarios.

Demonstrates effective high-level semantic encoding in watermarking.

Abstract

Rapid advances in generative AI have led to increasingly realistic deepfakes, posing growing challenges for law enforcement and public trust. Existing passive deepfake detectors struggle to keep pace, largely due to their dependence on specific forgery artifacts, which limits their ability to generalize to new deepfake types. Proactive deepfake detection using watermarks has emerged to address the challenge of identifying high-quality synthetic media. However, these methods often struggle to balance robustness against benign distortions with sensitivity to malicious tampering. This paper introduces a novel deep learning framework that harnesses high-dimensional latent space representations and the Multi-Agent Adversarial Reinforcement Learning (MAARL) paradigm to develop a robust and adaptive watermarking approach. Specifically, we develop a learnable watermark embedder that operates in the latent space, capturing high-level image semantics, while offering precise control over message encoding and extraction. The MAARL paradigm empowers the learnable watermarking agent to pursue an optimal balance between robustness and fragility by interacting with a dynamic curriculum of benign and malicious image manipulations simulated by an adversarial attacker agent. Comprehensive evaluations on the CelebA and CelebA-HQ benchmarks reveal that our method consistently outperforms state-of-the-art approaches, achieving improvements of over 4.5% on CelebA and more than 5.3% on CelebA-HQ under challenging manipulation scenarios.