Abstract
Automatic depression detection using audio-visual data faces significant challenges, particularly in disentangling overlapping feature distributions and establishing robust decision boundaries. To address this, we propose a fine-grained multimodal framework featuring a temporal encoder and a mutual transformer to facilitate deep cross-modal fusion.
Our core contribution is the Binary Advantage-weighted Ranking Loss, which optimizes the latent space distribution through two complementary mechanisms:
- Advantage-weighted Separation: This mechanism mines hard pairs by computing a pairwise prediction difference matrix and dynamically weighting them based on their difficulty.
- Advantage-weighted Compactness: This minimizes intra-class variance to force features to cluster around their respective class centers.
Extensive experiments on D-vlog and LMVD demonstrate that our model reconstructs the latent ordinal structure by prioritizing hard pairs, thereby achieving state-of-the-art performance.
Blogger's Review: This framework effectively addresses the feature overlap issue in depression detection, leveraging the advantages of multimodal data. It showcases the potential of deep learning in mental health applications, enhancing detection accuracy and providing new insights for future research.