A Dual-Branch CNN for Robust Detection of AI-Generated Facial Forgeries

2025-10-28

AI Summary

To address the digital trust crisis triggered by the proliferation of AI-generated facial forgery images, this paper proposes a dual-branch convolutional neural network that jointly models spatial- and frequency-domain features to enhance detection robustness. The frequency branch is innovatively designed to capture high-frequency artifacts inherently introduced by generative models, while a channel-wise attention mechanism enables adaptive fusion of multimodal features. Furthermore, we introduce the Frequency-Supervised Contrastive (FSC) unified loss function, integrating focal loss, supervised contrastive loss, and a novel frequency-center margin loss. Evaluated on the DiFF benchmark—comprising four forgery categories (text-to-image, image-to-image translation, face swapping, and attribute editing)—our method significantly outperforms existing state-of-the-art approaches, achieving detection accuracy surpassing human-level performance. The proposed framework establishes a new paradigm for AI-generated content detection, offering strong generalizability and enhanced interpretability.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

The rapid advancement of generative AI has enabled the creation of highly realistic forged facial images, posing significant threats to AI security, digital media integrity, and public trust. Face forgery techniques, ranging from face swapping and attribute editing to powerful diffusion-based image synthesis, are increasingly being used for malicious purposes such as misinformation, identity fraud, and defamation. This growing challenge underscores the urgent need for robust and generalizable face forgery detection methods as a critical component of AI security infrastructure. In this work, we propose a novel dual-branch convolutional neural network for face forgery detection that leverages complementary cues from both spatial and frequency domains. The RGB branch captures semantic information, while the frequency branch focuses on high-frequency artifacts that are difficult for generative models to suppress. A channel attention module is introduced to adaptively fuse these heterogeneous features, highlighting the most informative channels for forgery discrimination. To guide the network's learning process, we design a unified loss function, FSC Loss, that combines focal loss, supervised contrastive loss, and a frequency center margin loss to enhance class separability and robustness. We evaluate our model on the DiFF benchmark, which includes forged images generated from four representative methods: text-to-image, image-to-image, face swap, and face edit. Our method achieves strong performance across all categories and outperforms average human accuracy. These results demonstrate the model's effectiveness and its potential contribution to safeguarding AI ecosystems against visual forgery attacks.

Problem

• Detecting AI-generated facial forgeries to enhance security
• Combating malicious uses of face swapping and editing
• Improving robustness against visual forgery attacks

Innovation

• Dual-branch CNN combines spatial and frequency features
• Channel attention module adaptively fuses heterogeneous information
• Unified FSC Loss enhances class separability and robustness