La-SoftMoE CLIP for Unified Physical-Digital Face Attack Detection
This addresses a security risk in facial recognition systems by enabling detection of combined attacks, though it appears incremental as it builds on existing MoE and CLIP methods.
The paper tackles the problem of unified detection of physical and digital face attacks by proposing La-SoftMoE CLIP, which adapts the Mixture of Experts framework with a flexible weighting mechanism to handle sparse data and complex classification boundaries, achieving state-of-the-art performance.
Facial recognition systems are susceptible to both physical and digital attacks, posing significant security risks. Traditional approaches often treat these two attack types separately due to their distinct characteristics. Thus, when being combined attacked, almost all methods could not deal. Some studies attempt to combine the sparse data from both types of attacks into a single dataset and try to find a common feature space, which is often impractical due to the space is difficult to be found or even non-existent. To overcome these challenges, we propose a novel approach that uses the sparse model to handle sparse data, utilizing different parameter groups to process distinct regions of the sparse feature space. Specifically, we employ the Mixture of Experts (MoE) framework in our model, expert parameters are matched to tokens with varying weights during training and adaptively activated during testing. However, the traditional MoE struggles with the complex and irregular classification boundaries of this problem. Thus, we introduce a flexible self-adapting weighting mechanism, enabling the model to better fit and adapt. In this paper, we proposed La-SoftMoE CLIP, which allows for more flexible adaptation to the Unified Attack Detection (UAD) task, significantly enhancing the model's capability to handle diversity attacks. Experiment results demonstrate that our proposed method has SOTA performance.