DFCon: Attention-Driven Supervised Contrastive Learning for Robust Deepfake Detection

TL;DR

This paper introduces DFCon, a robust deepfake detection system using attention-driven supervised contrastive learning with multiple models and ensemble voting, achieving high accuracy across diverse datasets.

Contribution

The paper proposes a novel ensemble approach combining MaxViT, CoAtNet, and EVA-02 with supervised contrastive loss for improved deepfake detection in real-world scenarios.

Findings

Achieved 95.83% accuracy on validation dataset.

Effective combination of local and global feature extractors.

Ensemble voting enhances robustness and generalization.

Abstract

This report presents our approach for the IEEE SP Cup 2025: Deepfake Face Detection in the Wild (DFWild-Cup), focusing on detecting deepfakes across diverse datasets. Our methodology employs advanced backbone models, including MaxViT, CoAtNet, and EVA-02, fine-tuned using supervised contrastive loss to enhance feature separation. These models were specifically chosen for their complementary strengths. Integration of convolution layers and strided attention in MaxViT is well-suited for detecting local features. In contrast, hybrid use of convolution and attention mechanisms in CoAtNet effectively captures multi-scale features. Robust pretraining with masked image modeling of EVA-02 excels at capturing global features. After training, we freeze the parameters of these models and train the classification heads. Finally, a majority voting ensemble is employed to combine the predictions from these models, improving robustness and generalization to unseen scenarios. The proposed system addresses the challenges of detecting deepfakes in real-world conditions and achieves a commendable accuracy of 95.83% on the validation dataset.