CAMME: Adaptive Deepfake Image Detection with Multi-Modal Cross-Attention

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Khan, Naseem, Nguyen, Tuan, Bermak, Amine, Khalil, Issa
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913855798509568
author Khan, Naseem
Nguyen, Tuan
Bermak, Amine
Khalil, Issa
author_facet Khan, Naseem
Nguyen, Tuan
Bermak, Amine
Khalil, Issa
contents The proliferation of sophisticated AI-generated deepfakes poses critical challenges for digital media authentication and societal security. While existing detection methods perform well within specific generative domains, they exhibit significant performance degradation when applied to manipulations produced by unseen architectures--a fundamental limitation as generative technologies rapidly evolve. We propose CAMME (Cross-Attention Multi-Modal Embeddings), a framework that dynamically integrates visual, textual, and frequency-domain features through a multi-head cross-attention mechanism to establish robust cross-domain generalization. Extensive experiments demonstrate CAMME's superiority over state-of-the-art methods, yielding improvements of 12.56% on natural scenes and 13.25% on facial deepfakes. The framework demonstrates exceptional resilience, maintaining (over 91%) accuracy under natural image perturbations and achieving 89.01% and 96.14% accuracy against PGD and FGSM adversarial attacks, respectively. Our findings validate that integrating complementary modalities through cross-attention enables more effective decision boundary realignment for reliable deepfake detection across heterogeneous generative architectures.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18035
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CAMME: Adaptive Deepfake Image Detection with Multi-Modal Cross-Attention
Khan, Naseem
Nguyen, Tuan
Bermak, Amine
Khalil, Issa
Computer Vision and Pattern Recognition
F.2.2; I.2.7
The proliferation of sophisticated AI-generated deepfakes poses critical challenges for digital media authentication and societal security. While existing detection methods perform well within specific generative domains, they exhibit significant performance degradation when applied to manipulations produced by unseen architectures--a fundamental limitation as generative technologies rapidly evolve. We propose CAMME (Cross-Attention Multi-Modal Embeddings), a framework that dynamically integrates visual, textual, and frequency-domain features through a multi-head cross-attention mechanism to establish robust cross-domain generalization. Extensive experiments demonstrate CAMME's superiority over state-of-the-art methods, yielding improvements of 12.56% on natural scenes and 13.25% on facial deepfakes. The framework demonstrates exceptional resilience, maintaining (over 91%) accuracy under natural image perturbations and achieving 89.01% and 96.14% accuracy against PGD and FGSM adversarial attacks, respectively. Our findings validate that integrating complementary modalities through cross-attention enables more effective decision boundary realignment for reliable deepfake detection across heterogeneous generative architectures.
title CAMME: Adaptive Deepfake Image Detection with Multi-Modal Cross-Attention
topic Computer Vision and Pattern Recognition
F.2.2; I.2.7
url https://arxiv.org/abs/2505.18035