LAKAN: Landmark-assisted Adaptive Kolmogorov-Arnold Network for Face Forgery Detection

The rapid development of deepfake generation techniques necessitates robust face forgery detection algorithms. While methods based on Convolutional Neural Networks (CNNs) and Transformers are effective, there is still room for improvement in modeling the highly complex and non-linear nature of forgery artifacts. To address this issue, we propose a novel detection method based on the Kolmogorov-Arnold Network (KAN). By replacing fixed activation functions with learnable splines, our KAN-based approach is better suited to this challenge. Furthermore, to guide the network's focus towards critical facial areas, we introduce a Landmark-assisted Adaptive Kolmogorov-Arnold Network (LAKAN) module. This module uses facial landmarks as a structural prior to dynamically generate the internal parameters of the KAN, creating an instance-specific signal that steers a general-purpose image encoder towards the most informative facial regions with artifacts. This core innovation creates a powerful combination between geometric priors and the network's learning process. Extensive experiments on multiple public datasets show that our proposed method achieves superior performance.

Key Contributions

• Introduces KAN (Kolmogorov-Arnold Network) with learnable B-spline activations for face forgery detection, offering superior non-linear fitting over fixed-activation CNNs/Transformers
• Proposes the LAKAN module, which uses facial landmarks as structural priors to dynamically and instance-specifically generate KAN's internal spline parameters
• Applies LAKAN as an adaptive gating mechanism across multiple stages of a general-purpose image encoder to steer attention toward forgery-rich facial regions

🛡️ Threat Analysis

Details

Similar Papers