Pith. sign in

REVIEW 3 major objections 4 minor 48 references

Large Language Models for Depression Recognition in Spoken Language Integrating Psychological Knowledge

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that an LLM pipeline fusing Wav2Vec audio embeddings with transcripts and post-trained on psychology Q&A pairs estimates PHQ-8 depression severity on DAIC-WOZ more accurately than the AVEC2016 audio baseline.

desk verdict Table 2 as printed contradicts the paper's headline: the full multimodal pipeline is identical to audio-only on MAE/RMSE, so the claimed benefit isn't there. read the letter →

arxiv 2505.22863 v1 pith:XS2XTSIO submitted 2025-05-28 cs.HC cs.CL

Yupei Li , Shuaijie Shao , Manuel Milling , Björn W. Schuller This is my paper
classification cs.HCcs.CL
keywords depressionrecognitionlargelanguagemodelsmultimodalfusionpsychologicalknowledgeinjectionDAIC-WOZPHQ-8Wav2Vec2.0speech-basedLLM
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that a large language model can be turned into a spoken-language depression severity estimator by combining audio and text. It projects Wav2Vec 2.0 audio features into LLaMA's embedding space, adds interview transcripts, and first post-trains the model on thousands of question-answer pairs distilled from clinical classification entries. On the DAIC-WOZ test set the complete pipeline reaches MAE 5.356 and RMSE 6.713, below the AVEC2016 audio baseline's 5.72 and 7.78. The authors also report that the psychology-knowledge injection improves the text-only model from MAE 6.342 to 5.354 and RMSE 8.891 to 6.429. A sympathetic reader would take the paper as evidence that knowledge-injected LLMs are a viable route for multimodal mental-health assessment, not as a deployed diagnostic.

What carries the argument

The load-bearing machinery is a two-stage fine-tuning protocol on LLaMA. First, a knowledge-injection stage trains the model with supervised learning to answer 4,920 DeepSeek-generated question-answer pairs about depression drawn from clinical classification entries, using question types such as definition, rationale, symptoms, extended knowledge, and critical thinking. Second, a multi-stream stage feeds Wav2Vec 2.0 embeddings through a feed-forward projector to align them with LLaMA's text embedding space, trains the model on transcript prompts of the form 'Transcripts:..., PHQ Score:' and on projected audio, then applies a linear head to the last hidden state to output a PHQ-8 score; LoRA keeps the fine-tuning cheap. This two-stage design is what lets a text-native LLM consume acoustic cues and domain knowledge at once.

What would settle it

Compute the within-participant variance of the model's clip-level PHQ-8 predictions: if predictions are almost constant inside a participant and vary mainly across participants, that is direct evidence the reported MAE/RMSE comes from person-level label repetition rather than from detecting depression in each clip.

Watch

Extended reading notes

Core claim

The central discovery claimed is that a two-stage pipeline—Wav2Vec 2.0 for audio, a feed-forward projector, LLaMA as backbone, and a supervised post-training stage on 4,920 Q&A pairs generated from clinical depression entries—regresses PHQ-8 scores at the segment level and averages segment predictions per participant, beating the AVEC2016 audio baseline by about 0.36 MAE and 1.07 RMSE. The comparison includes ablations: audio-only, text-only, audio+text, each with and without knowledge injection. According to the paper, knowledge injection helps the text stream strongly and helps the audio stream slightly, while adding audio to text plus knowledge does not substantially change the numbers further.

Load-bearing premise

The evaluation assumes that every five-utterance clip from a participant is equally well described by that participant's single whole-interview depression score, so the model is trained and scored against duplicated participant-level labels without evidence that segments are homogeneous.

Editorial extensions

If this is right

  • If the result holds, a text-native LLM can serve as the fusion backbone for acoustic and lexical depression cues without a dedicated speech-language model.
  • Injecting structured psychological knowledge before task fine-tuning is a transferable way to improve LLM-based mental-health scoring, with the larger effect on text than on audio.
  • The reported margin over the AVEC2016 audio baseline is 0.364 MAE and 1.067 RMSE for the full model, so the practical value lies in the pipeline design rather than in a large accuracy jump.
  • The near-equal scores for audio+text+knowledge, audio+knowledge, and audio+text suggest that with LoRA-scale tuning the audio stream dominates or saturates the fusion, making larger-capacity training the natural next step named by the authors.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, the same knowledge-injection protocol could be tested on per-utterance or per-visit depression labels to see whether the model captures within-interview symptom dynamics rather than only person-level severity.
  • A direct extension would be to feed the model's clip-level predictions into a clinical decision-support display: if the averaging step hides high-variance segments, flagging segments with extreme predicted scores might give clinicians more useful signals than the averaged participant number.
  • One consequence the authors do not develop is that their Q&A-generation step could be repeated for other disorders in the same classification source, turning the pipeline into a general knowledge-injected health LLM rather than a depression-specific one.
  • Because the knowledge injection is text-only, an audio-description variant—pairing each clip with a text caption of its prosody, pauses, and pitch variability—would test whether the transfer from text knowledge to audio can be strengthened, which the authors list as future work.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The manuscript proposes a multimodal pipeline for depression severity regression on the DAIC-WOZ dataset, combining Wav2Vec 2.0 audio embeddings, ASR transcripts, and a psychological knowledge-injection step in which DeepSeek generates question–answer pairs from WHO texts and LLaMA is fine-tuned on them with LoRA. The authors report MAE and RMSE for several ablations, compare them with AVEC 2016 audio and LSTM/Random Forest baselines, and claim notable improvements in both metrics. The paper also includes a self-assessment of the knowledge-injection step using PsycholexLLAMA as both question generator and answer grader.

Significance. If the reported results were valid, the paper would offer a useful practical contribution: a reproducible, code-released LLM-based approach to multimodal depression recognition with a mechanism for injecting domain knowledge. The authors also deserve credit for fixing the random seed, reporting hyperparameters, and performing a structured ablation. However, the central empirical claim is not supported by the manuscript as written. Table 2 shows that the full multimodal model with knowledge injection achieves exactly the same rounded MAE (5.356) and RMSE (6.713) as the audio-only and audio-plus-knowledge variants, so the text stream and the knowledge injection have no measurable effect when audio is present. The evaluation protocol also assigns the same per-participant PHQ-8 score to every clip from that participant, which leaves open the possibility that the model is exploiting participant identity rather than depression-relevant content. These are load-bearing problems: the abstract's 'notable improvement' claim, the conclusion's numerical summary, and the claimed benefit of psychological knowledge injection all rest on numbers that the paper itself does not consistently support.

major comments (3)
  1. [§3, Table 2; §4 (Conclusion)] The central empirical claim is contradicted by Table 2 as printed. 'Ours(audio+text+Knowledge Injection)' reports MAE 5.356 and RMSE 6.713, identical at the reported precision to both 'Ours(audio+text)' and 'Ours(audio+Knowledge Injection)'; consequently, the results show no effect of the text stream and no effect of knowledge injection when audio is present. The full model's RMSE of 6.713 is also worse than the LSTM baseline of 6.59 listed in the same table. In addition, the conclusion's stated improvement of '1.36 on RMSE' corresponds to comparing the AVEC2016 baseline (7.78) with 'Ours(text+Knowledge Injection)' (6.429), not with the full multimodal model, whose RMSE improvement over the AVEC2016 baseline is only 7.78 − 6.713 = 1.067. The authors must correct the table or rerun the experiments; as printed, neither the abstract's 'notable improvement' nor the conclusion's summary is supported by the data.
  2. [§2.2.2 (Data Preprocessing)] The evaluation protocol assigns a single participant-level PHQ-8 score to every five-utterance clip from that participant and then averages clip-level predictions back to a participant score. Because all clips from one participant share the same target, the model could in principle learn to recognize participant identity or interview-level acoustic properties rather than depression-relevant content within each clip. The manuscript provides no analysis of within-participant label homogeneity, no participant-wise cross-validation, and no baseline that controls for participant identity. This is load-bearing because the headline MAE/RMSE values depend on this clip-to-participant aggregation, and the reported numbers therefore cannot be interpreted as evidence of depression detection per segment.
  3. [§3, knowledge-injection quality check] The quality check for knowledge injection is self-referential: PsycholexLLAMA generates the fifty questions, and the same model then scores answers from the original and trained models, with a pass@2-style maximum taken. There is no human evaluation or external benchmark. Because the trained model was fine-tuned on the same style of DeepSeek-generated Q&A pairs, its higher average score (8.20 vs. 7.32) may reflect stylistic conformity or in-domain self-preference rather than improved psychological knowledge. This check does not by itself validate the claim that the injected knowledge improves depression scoring.
minor comments (4)
  1. [§2.3.1 (Psychology Knowledge Injection)] The text says the authors design 'six distinct types' of questions, but the enumerated list contains only five items: definition, rationale, symptoms, extended knowledge, and critical thinking. The count should be corrected or the sixth type should be stated explicitly.
  2. [Abstract and §1] The claim of being 'the first application of LLMs to multimodal depression detection using the DAIC-WOZ dataset' is stated too strongly. The cited related work by Sadeghi et al. (2024) uses an LLM combined with facial-expression features for PHQ-8 regression, and Tank et al. (2024) applies LLMs to textual and audio-visual modalities; the novelty claim needs qualification in light of these and similar prior works.
  3. [§2.3.2, Eq. (3)] Equation (3) writes 'PHQ-8 = Linear(LLaMA(Embaudio)−1)', but the superscript '−1' is unexplained and appears to be a typographical artifact. Please clarify whether it denotes the last hidden layer or something else.
  4. [Throughout] There are several minor presentation issues: the duplicated sentence about lexicon-based sentiment indicators in §2.1.1, irregular spacing in Table 2 row labels such as 'Ours(audio +text+ Knowledge Injection)', and the vague phrase 'mapped it to text-based LLMs' in the abstract. These should be cleaned up in a revision.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity in the central derivation; reported internal inconsistencies are correctness issues, not circularity.

full rationale

The paper's central claim is that a multimodal LLaMA pipeline with Wav2Vec audio features and WHO-derived knowledge injection improves PHQ-8 regression. This claim is anchored to external DAIC-WOZ PHQ-8 labels and to the AVEC 2016 audio baseline; the audio, text, and knowledge-injection inputs are not constructed from the predicted scores. Wav2Vec 2.0 and LLaMA are externally pretrained, and the LoRA fine-tuning minimizes an external regression and answer-generation loss. Section 2.2.2's decision to assign each participant's single PHQ-8 score to every five-utterance clip is label duplication, but the final participant score is an average of clip predictions, so it does not make the prediction equal to the label by construction. The self-referential check in Section 3, where PsycholexLLAMA both generates and grades the questions and the pass@2 max is attributed to the trained model, is a weak and potentially misleading validation, but it is not the load-bearing derivation of the DAIC-WOZ numbers and no equation or fitted parameter reduces to its own input. The duplicated rows in Table 2 (audio+text = audio+knowledge injection = full model) and the omission of the LSTM and Random Forest comparisons from the headline are internal reporting inconsistencies that should be corrected, but they are correctness issues rather than circularity. Under the requirement to exhibit a specific by-construction reduction, no such reduction is present.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central result depends on two hand-chosen hyperparameters and four unvalidated domain assumptions. No new entities are introduced. The largest burdens are the per-clip label assignment and the unvalidated premise that DeepSeek-generated WHO Q&A pairs act as authoritative clinical knowledge.

free parameters (2)
  • Utterances merged per clip = 5
    Section 2.2.2 chooses every five consecutive utterances as a segment. No ablation over this window size is reported, yet it determines the granularity of every prediction and the number of clips (6,556).
  • LoRA hyperparameters = r=8, alpha=16, dropout=0.1
    Table 1 lists these as fixed hyperparameters with no tuning or sensitivity analysis; the paper itself notes LoRA's limited capacity to change model behavior.
assumptions (4)
  • domain assumption Depression severity is reflected in audio features extractable by Wav2Vec 2.0
    The whole pipeline rests on the premise that vocal tone, prosody, and other acoustic cues carry depression information that a self-supervised speech model can capture; Section 2.3.2 invokes Wav2Vec 2.0 for 'emotional cues and prosodic elements associated with depression.'
  • domain assumption A participant-level PHQ-8 score is a valid label for every five-utterance clip
    Section 2.2.2 assigns the same PHQ-8 score to all clips from a speaker; if clip-level depression expression varies, the labels are noisy duplicates of one score.
  • ad hoc to paper DeepSeek-generated Q&A pairs from WHO text constitute authoritative psychological knowledge whose supervised fine-tuning improves depression scoring
    Section 2.3.1 builds 4,920 pairs by prompting DeepSeek with WHO ICD entries; there is no human clinician check that these pairs are correct or complete, and no evidence that this injection transfers to audio regression.
  • ad hoc to paper PsycholexLLAMA's 0-10 ratings are a valid measure of answer quality
    Section 3 uses the same model family to generate questions and score answers, with no human validation, so the 8.20 vs 7.32 result is a self-assessment.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Large Language Models for Depression Recognition in Spoken Language Integrating Psychological Knowledge." pith.science (2026). https://pith.science/paper/XS2XTSIO

@misc{pith2026250522863,
  author       = {Pith},
  title        = {Pith review of: Large Language Models for Depression Recognition in Spoken Language Integrating Psychological Knowledge},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XS2XTSIO}},
  note         = {Machine review of arXiv:2505.22863}
}
read the original abstract

Depression is a growing concern gaining attention in both public discourse and AI research. While deep neural networks (DNNs) have been used for recognition, they still lack real-world effectiveness. Large language models (LLMs) show strong potential but require domain-specific fine-tuning and struggle with non-textual cues. Since depression is often expressed through vocal tone and behaviour rather than explicit text, relying on language alone is insufficient. Diagnostic accuracy also suffers without incorporating psychological expertise. To address these limitations, we present, to the best of our knowledge, the first application of LLMs to multimodal depression detection using the DAIC-WOZ dataset. We extract the audio features using the pre-trained model Wav2Vec, and mapped it to text-based LLMs for further processing. We also propose a novel strategy for incorporating psychological knowledge into LLMs to enhance diagnostic performance, specifically using a question and answer set to grant authorised knowledge to LLMs. Our approach yields a notable improvement in both Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) compared to a base score proposed by the related original paper. The codes are available at https://github.com/myxp-lyp/Depression-detection.git

Figures

Figures reproduced from arXiv: 2505.22863 by the authors.

Figure 1
Figure 1. Two-Stage Pipeline of the Large Models considered. Our proposed framework consists of two key stages. In the first stage, we leverage the DeepSeek model to extract question–answer pairs from authoritative psychology texts, such as disease definitions and clinical descriptions from the World Health Organization (WHO). These extracted pairs, along with transcript data, are used to pretrain the LLM through a process of… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

48 extracted references · 34 canonical work pages

  1. [1]

    PsychoLex: Unveiling the Psychological Mind of Large Language Models

    Abbasi, M. A., Mirnezami, F. S., and Naderi, H. (2024). Psycholex: Unveiling the psychological mind of large language models. arXiv preprint arXiv:2408.08848 abbasi2024psycholexunveilingpsychologicalmind

    work page Pith review arXiv 2024

  2. [2]

    and Khodaei, N

    Afzal Aghaei, A. and Khodaei, N. (2023). Automated depression recognition using multimodal machine learning: A study on the daic-woz dataset. Computational Mathematics and Computer Modeling with Applications (CMCMA) 2, 45--53 afzal2023automated

    work page 2023

  3. [3]

    M., Cambria, E., and Schuller, B

    Amin, M. M., Cambria, E., and Schuller, B. W. (2023). Will affective computing emerge from foundation models and general artificial intelligence? a first evaluation of chatgpt. IEEE Intelligent Systems 38, 15--23. doi:10.1109/MIS.2023.3254179 10111523

    arXiv 2023

  4. [4]

    S., Riza, B

    Ashraf, A., Gunawan, T. S., Riza, B. S., Haryanto, E. V., and Janin, Z. (2020). On the review of image and video-based depression detection using machine learning. Indonesian Journal of Electrical Engineering and Computer Science (IJEECS) 19, 1677--1684 ashraf2020review

    work page 2020

  5. [5]

    Baevski, A., Zhou, Y., Mohamed, A., and Auli, M. (2020). wav2vec 2.0: A framework for self-supervised learning of speech representations. Advances in neural information processing systems 33, 12449--12460 baevski2020wav2vec20frameworkselfsupervised

    work page 2020

  6. [6]

    Berardi, M., Brosch, K., Pfarr, J.-K., Schneider, K., S \"u ltmann, A., Thomas-Odenthal, F., et al. (2023). Relative importance of speech and voice features in the classification of schizophrenia and depression. Translational psychiatry 13, 298 berardi2023relative

    work page 2023

  7. [7]

    Chen, M., Tworek, J., Jun, H., Yuan, Q., Pinto, H. P. D. O., Kaplan, J., et al. (2021). Evaluating large language models trained on code. arXiv preprint arXiv:2107.03374 chen2021evaluating

    arXiv 2021

  8. [8]

    Farruque, N., Goebel, R., Sivapalan, S., and Za \" ane, O. R. (2024). Depression symptoms modelling from social media text: an llm driven semi-supervised learning approach. Language Resources and Evaluation 58, 1013--1041 farruque2024depression

    work page 2024

Show all 48 references
  1. [9]

    gold standard

    Gerber, P. D., Barrett, J., Barrett, J., Manheimer, E., Whiting, R., and Smith, R. (1989). Recognition of depression by internists in primary care: a comparison of internist and “gold standard” psychiatric assessments. Journal of General Internal Medicine 4, 7--13 gerber1989re...

    1989

  2. [10]

    Gratch, J., Artstein, R., Lucas, G., Stratou, G., Scherer, S., Nazarian, A., et al. (2014). The distress analysis interview corpus of human and computer interviews. In Proceedings of the Ninth International Conference on Language Resources and Evaluation ( LREC `14) , eds. N. ...

    2014

  3. [11]

    T., et al

    He, X., Triantafyllopoulos, A., Kathan, A., Milling, M., Yan, T., Rajamani, S. T., et al. (2022). Depression diagnosis and forecast based on mobile phone sensor data. In 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC) (IEE...

    2022

  4. [12]

    J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., et al

    Hu, E. J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., et al. (2022). Lora: Low-rank adaptation of large language models. ICLR 1, 3 hu2021loralowrankadaptationlarge

    2022

  5. [13]

    Huang, X., Wang, F., Gao, Y., Liao, Y., Zhang, W., Zhang, L., et al. (2024). Depression recognition using voice-based pre-training model. Scientific Reports 14, 12734 huang2024depression

    2024

  6. [14]

    (2022 a )

    Kathan, A., Harrer, M., K \"u ster, L., Triantafyllopoulos, A., He, X., Milling, M., et al. (2022 a ). Personalised depression forecasting using mobile sensor data and ecological momentary assessment. Frontiers in digital health 4, 964582 kathan2022personalised

    2022

  7. [15]

    T., et al

    Kathan, A., Triantafyllopoulos, A., He, X., Milling, M., Yan, T., Rajamani, S. T., et al. (2022 b ). Journaling data for daily phq-2 depression prediction and forecasting. In 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC)...

    2022

  8. [16]

    Lan, X., Cheng, Y., Sheng, L., Gao, C., and Li, Y. (2024). Depression detection on social media with large language models. arXiv preprint arXiv:2403.10750 lan2024depression

    2024

  9. [17]

    Li, Y., Milling, M., and Schuller, B. W. (2025). Neuroplasticity in artificial intelligence--an overview and inspirations on drop in & out learning. arXiv preprint arXiv:2503.21419 li2025neuroplasticityartificialintelligence

    2025 arXiv

  10. [18]

    (2024 a )

    Liu, A., Feng, B., Xue, B., Wang, B., Wu, B., Lu, C., et al. (2024 a ). Deepseek-v3 technical report. arXiv preprint arXiv:2412.19437 liu2024deepseek

    2024 arXiv

  11. [19]

    (2024 b )

    Liu, Z., Yang, K., Xie, Q., Zhang, T., and Ananiadou, S. (2024 b ). Emollms: A series of emotional large language models and annotation tools for comprehensive affective analysis. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 5487--54...

    2024

  12. [20]

    Lyu, S., Ren, X., Du, Y., and Zhao, N. (2023). Detecting depression of chinese microblog users via text analysis: Combining linguistic inquiry word count (liwc) with culture and suicide related lexicons. Frontiers in Psychiatry Volume 14 - 2023. doi:10.3389/fpsyt.2023.1121583 ...

    2023

  13. [21]

    Mallol-Ragolta, A., Milling, M., and Schuller, B. (2024). Multi-triplet loss-based models for categorical depression recognition from speech. In Proc. IberSPEECH 2024. 31--35 mallol2024multi

    2024

  14. [22]

    Martino, A., Iannelli, M., and Truong, C. (2023). Knowledge injection to counter large language model (llm) hallucination. In European Semantic Web Conference (Springer), 182--185 martino2023knowledge

    2023

  15. [23]

    Min, K., Yoon, J., Kang, M., Lee, D., Park, E., and Han, J. (2023). Detecting depression on video logs using audiovisual features. Humanities and Social Sciences Communications 10, 1--8 min2023detecting

    2023

  16. [24]

    Niu, M., Zhao, Z., Tao, J., Li, Y., and Schuller, B. W. (2022). Dual attention and element recalibration networks for automatic depression level prediction. IEEE Transactions on Affective Computing 14, 1954--1965 niu2022dual

    2022

  17. [25]

    Novak, J. D. and Gowin, D. B. (1984). Learning how to learn (cambridge University press) novak1984learning

    1984

  18. [26]

    Ogunleye, B., Sharma, H., and Shobayo, O. (2024). Sentiment informed sentence bert-ensemble algorithm for depression detection. Big Data and Cognitive Computing 8, 112 ogunleye2024sentiment

    2024

  19. [27]

    Qassim, S., Golden, G., Slowey, D., Sarfas, M., Whitmore, K., Perez, T., et al. (2023). A mixed-methods feasibility study of a novel ai-enabled, web-based, clinical decision support system for the treatment of major depression in adults. Journal of Affective Disorders Reports ...

    2023

  20. [28]

    W., Xu, T., Brockman, G., McLeavey, C., and Sutskever, I

    Radford, A., Kim, J. W., Xu, T., Brockman, G., McLeavey, C., and Sutskever, I. (2023). Robust speech recognition via large-scale weak supervision. In International conference on machine learning (PMLR), 28492--28518 radford2022robustspeechrecognitionlargescale

    2023

  21. [29]

    Ringeval, F., Schuller, B., Valstar, M., Cummins, N., Cowie, R., Tavabi, L., et al. (2019). Avec 2019 workshop and challenge: state-of-mind, detecting depression with ai, and cross-cultural affect recognition. In Proceedings of the 9th International on Audio/visual Emotion Cha...

    2019

  22. [30]

    Ringeval, F., Schuller, B., Valstar, M., Gratch, J., Cowie, R., Scherer, S., et al. (2017). Avec 2017: Real-life depression, and affect recognition workshop and challenge. In Proceedings of the 7th Annual Workshop on Audio/Visual Emotion Challenge (New York, NY, USA: Associati...

    2017

  23. [31]

    Robertson, C., Woods, A., Bergstrand, K., Findley, J., Balser, C., and Slepian, M. J. (2023). Diverse patients’ attitudes towards artificial intelligence (ai) in diagnosis. PLOS Digital Health 2, e0000237 robertson2023diverse

    2023

  24. [32]

    M., and Durairaj, T

    S, S., V, S., N, S., C, J. M., and Durairaj, T. (2022). scube MSEC @ LT - EDI - ACL 2022: Detection of depression using transformer models. In Proceedings of the Second Workshop on Language Technology for Equality, Diversity and Inclusion, eds. B. R. Chakravarthi, B. Bharathi,...

    2022 doi

  25. [33]

    H., Rahimi, F., et al

    Sadeghi, M., Richer, R., Egger, B., Schindler-Gmelch, L., Rupp, L. H., Rahimi, F., et al. (2024). Harnessing multimodal approaches for depression detection using large language models and facial expressions. npj Mental Health Research 3, 66 sadeghi2024harnessing

    2024

  26. [34]

    N., and Eklund, P

    Sardari, S., Nakisa, B., Rastgoo, M. N., and Eklund, P. (2022). Audio based depression detection using convolutional autoencoder. Expert Systems with Applications 189, 116076 sardari2022audio

    2022

  27. [35]

    P., Tsangko, I., Amin, M

    Schuller, B., Mallol-Ragolta, A., Almansa, A. P., Tsangko, I., Amin, M. M., Semertzidou, A., et al. (2024). Affective computing has changed: The foundation model disruption. arXiv preprint arXiv:2409.08907 schuller2024affective

    2024 arXiv

  28. [36]

    Shin, D., Kim, H., Lee, S., Cho, Y., and Jung, W. (2024). Using large language models to detect depression from user-generated diary text data as a novel approach in digital mental health screening: Instrument validation study. Journal of Medical Internet Research 26, e54617 s...

    2024

  29. [37]

    R., et al

    Squires, M., Tao, X., Elangovan, S., Gururajan, R., Zhou, X., Acharya, U. R., et al. (2023). Deep learning and machine learning in psychiatry: a survey of current progress in depression detection, diagnosis and treatment. Brain Informatics 10, 10 squires2023deep

    2023

  30. [38]

    Tang, C., Yu, W., Sun, G., Chen, X., Tan, T., Li, W., et al. (2023). Salmonn: Towards generic hearing abilities for large language models. arXiv preprint arXiv:2310.13289 tang2024salmonngenerichearingabilities

    2023 arXiv

  31. [39]

    Tank, C., Pol, S., Katoch, V., Mehta, S., Anand, A., and Shah, R. R. (2024). Depression detection and analysis using large language models on textual and audio-visual modalities. arXiv preprint arXiv:2407.06125 tank2024depression

    2024 arXiv

  32. [40]

    Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.-A., Lacroix, T., et al. (2023). Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971 touvron2023llamaopenefficientfoundation

    2023 arXiv

  33. [41]

    Valstar, M., Gratch, J., Schuller, B., Ringeval, F., Lalanne, D., Torres Torres, M., et al. (2016). Avec 2016: Depression, mood, and emotion recognition workshop and challenge. In Proceedings of the 6th international workshop on audio/visual emotion challenge. 3--10 valstar201...

    2016

  34. [42]

    Valstar, M., Schuller, B., Smith, K., Eyben, F., Jiang, B., Bilakhia, S., et al. (2013). Avec 2013: the continuous audio/visual emotion and depression recognition challenge. In Proceedings of the 3rd ACM international workshop on Audio/visual emotion challenge. 3--10 valstar2013avec

    2013

  35. [43]

    F., Vivas, R

    Zafar, F., Alam, L. F., Vivas, R. R., Wang, J., Whei, S. J., Mehmood, S., et al. (2024). The role of artificial intelligence in identifying depression and anxiety: a comprehensive literature review. Cureus 16 zafar2024role

    2024

  36. [44]

    Zhang, Z., Peng, L., Pang, T., Han, J., Zhao, H., and Schuller, B. W. (2024). Refashioning emotion recognition modelling: The advent of generalised large models. IEEE Transactions on Computational Social Systems zhang2023refashioningemotionrecognitionmodelling

    2024

  37. [45]

    , " * write output.state after.block = add.period write newline

    ENTRY address annote author booktitle chapter doi edition editor eid howpublished institution journal key language month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.s...

  38. [46]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

  39. [47]

    , " * write output.state after.block = add.period write newline

    ENTRY address author booktitle chapter doi edition editor eid howpublished institution journal key language month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence...

  40. [48]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.