Fine-Grained DINO Tuning with Dual Supervision for Face Forgery Detection
This work addresses the threat of deepfakes to information integrity by improving detection accuracy in a parameter-efficient way, though it is incremental as it builds on existing DINOv2 fine-tuning approaches.
The paper tackled the problem of detecting deepfakes by fine-tuning DINOv2 with a novel adapter that addresses both authenticity detection and manipulation type classification, achieving detection accuracy comparable to or surpassing state-of-the-art methods using only 3.5M trainable parameters.
The proliferation of sophisticated deepfakes poses significant threats to information integrity. While DINOv2 shows promise for detection, existing fine-tuning approaches treat it as generic binary classification, overlooking distinct artifacts inherent to different deepfake methods. To address this, we propose a DeepFake Fine-Grained Adapter (DFF-Adapter) for DINOv2. Our method incorporates lightweight multi-head LoRA modules into every transformer block, enabling efficient backbone adaptation. DFF-Adapter simultaneously addresses authenticity detection and fine-grained manipulation type classification, where classifying forgery methods enhances artifact sensitivity. We introduce a shared branch propagating fine-grained manipulation cues to the authenticity head. This enables multi-task cooperative optimization, explicitly enhancing authenticity discrimination with manipulation-specific knowledge. Utilizing only 3.5M trainable parameters, our parameter-efficient approach achieves detection accuracy comparable to or even surpassing that of current complex state-of-the-art methods.