Depression manifests through a diverse set of symptoms such as sleep disturbance, loss of interest, and concentration difficulties. However, most existing works treat depression prediction either as a binary label or an overall severity score without explicitly modeling symptom-specific information. This limits their ability to provide symptom-level analysis relevant to clinical screening. To address this, we propose a symptom-specific and clinically inspired framework for depression severity estimation from speech. Our approach uses a symptom-guided cross-attention mechanism that aligns PHQ-8 questionnaire items with emotion-aware speech representations to identify which segments of a participant's speech are more important to each symptom. To account for differences in how symptoms are expressed over time, we introduce a learnable symptom-specific parameter that adaptively controls the sharpness of attention distributions. Our results on EDAIC, a standard clinical-style dataset, demonstrate improved performance outperforming prior works. Further, analyzing the attention distributions showed that higher attention is assigned to utterances containing cues related to multiple depressive symptoms, highlighting the interpretability of our approach. These findings outline the importance of symptom-guided and emotion-aware modeling for speech-based depression screening.
Figures & tables
Fig. 1 : Proposed approach inspired from PHQ-8 based clinial practice
Fig. 2 : Block diagram of the overall framework
Method
RMSE
MAE
CCC
Zhang et al. [ 19 ]
6.78
5.77
–
Fan et al. [ 17 ]
5.91
4.39
0.43
Yin et al. [ 20 ]
5.50
–
0.44
Zhao et al. [ 21 ]
5.78
5.13
–
Makiuchi et al. [ 22 ]
6.11
–
0.40
Han et al. [ 23 ]
6.29
5.38
–
TABLE I : Comparison with prior works on EDAIC
Symptom
RMSE
MAE
No Interest
0.891
0.638
Depressed
0.728
0.571
Sleep
1.104
0.903
Tired
1.005
0.842
Appetite
1.008
0.800
Failure
0.903
0.734
TABLE II : Sypmtom-wise Performance Metrics
τ setting
RMSE
MAE
CCC
No τ
5.36
4.22
0.55
Learnable τ (global)
5.31
4.17
0.54
Learnable τ ( per symptom)
5.15
4.13
0.52
TABLE III : Performance comparison with different τ settings
Fig. 3 : Visualization of attention heatmap and text alignment of the sentences marked with ’x’
Automatic depression detection using speech signals with acoustic and textual modalities is a promising approach for early diagnosis. Depression-related patterns exhibit sparsity in speech: diagnostically relevant features occur in specific segments rather than being uniformly distributed. However, most existing methods treat all frames equally, assuming depression-related information is uniformly distributed and thus overlooking this sparsity. To address this issue, we proposes a depression detection network based on Adaptive Cross-Modal Gating (ACMG) that adaptively reassigns frame-level weights across both modalities, enabling selective attention to depression-related segments. Experimental results show that the depression detection system with ACMG outperforms baselines without it. Visualization analyses further confirm that ACMG automatically attends to clinically meaningful patterns, including low-energy acoustic segments and textual segments containing negative sentiments.
Hangbin Yu, Yudong Yang, Rongfeng Su +2
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, China · University of Chinese Academy of Sciences · Key Laboratory of Biomedical Imaging Science and System, Chinese Academy of Sciences, China
Text-based counseling provides a valuable source of information for assessing depression severity. We study prediction of the total score on the eight-item Patient Health Questionnaire (PHQ-8), a self-report measure of depression severity, from counseling transcripts. Clinical-based methods rely mainly on large language model (LLM) inference to obtain structured session-level assessments, but these assessments provide limited information about which utterances support each score. Training-based methods train predictors directly on sentences or their semantic embeddings and preserve local conversational detail, but must learn clinical structure from limited labeled transcripts. Integrating a small set of clinical feature scores with a long sequence of high-dimensional utterance representations is challenging, as simple fusion can overemphasize dialogue evidence and fail to link clinical features to supporting utterances. Combining structured clinical assessments with sentence-level semantics, we propose EmoTrack, which jointly encodes clinical feature scores and utterance representations to predict the total PHQ-8 score. For longitudinal assessment across successive sessions, the model can also incorporate a compressed representation of the preceding session as optional memory. On the real-world DAIC-WOZ benchmark, EmoTrack reduces mean absolute error from 2.8234 to 2.4708 relative to the strongest evaluated baseline. Across five settings covering single-session estimation, cross-dataset transfer without adaptation, and longitudinal assessment, it reduces normalized aggregate error by approximately 6.1% relative to the strongest baseline.
Zhaomin Wu, Jiayi Li, Bingsheng He
Department of Computer Science National University of Singapore
Automatic depression detection from conversational interactions holds significant promise for scalable screening but remains hindered by severe data scarcity and a lack of clinical interpretability. Existing approaches typically rely on black-box deep learning architectures that struggle to model the subtle, temporal evolution of depressive symptoms or account for participant-specific heterogeneity. In this work, we propose PsyGAT (Psychological Graph Attention Network), a psychologically grounded framework that models conversational sessions as dynamic temporal graphs. We introduce Psychological Expression Units (PEUs) to explicitly encode utterance-level clinical evidence, structuring the session graph to capture transitions in psychological states rather than mere semantic dependencies. To address the critical class imbalance in depression datasets, we employ clinically approved persona-based data augmentation, enable robust model learning. Additionally, we integrate session-level personality context directly into the graph structure to disentangle trait-based behavior from acute depressive symptoms. PsyGAT achieves state-of-the-art performance, surpassing both strong graph-based baselines and closed-source LLMs like GPT-5, achieving 89.99 and 71.37 Macro F1 scores in DAIC-WoZ and E-DAIC, respectively. We further introduce Causal-PsyGAT, an interpretability module that identifies symptom triggers. Experiments show a 20% improvement in MRR for identifying causal indicators, effectively bridging the gap between depression monitoring and clinical explainability. The full augmented dataset is publicly available at https://doi.org/10.6084/m9.figshare.31801921.