ModernBert Outperforms Larger Models in Generalizing Social Media Depression Detection
A benchmark study reveals ModernBert outperforms larger language models on unseen social media data for depression detection.
A benchmark study reveals ModernBert outperforms larger language models on unseen social media data for depression detection.
A leakage-controlled multimodal framework combining speech, facial landmarks, and text embeddings achieved 91.2 percent accuracy in identifying depression on the ...
A narrative review reveals that artificial intelligence tools for detecting depression show marked performance declines when evaluated in large cohorts ...
A federated, differentially private multimodal framework detects depression and anxiety across three benchmarks with F1 up to 0.806 while cutting ...
Researchers report a Mamba-BiLSTM framework with Granger-inspired directional causal fusion that reached 81.86 percent accuracy for multimodal depression detection on ...
A new deep learning framework that passively reads smartphone sensor data from college students can detect depression and stress, but ...
A new CNN-Transformer model detects depression from Chinese speech with over 95% precision by fusing wavelet-based spectral features with hybrid ...
© 2026 Scienmag - Science Magazine
© 2026 Scienmag - Science Magazine