EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

StressSpeak: A Speech-Driven Framework for Real-Time Personalized Stress Detection and Adaptive Psychological Support

Laraib Umer, Javaid Iqbal, Yasar Ayaz iD, Hassan Imam iD, Adil Ahmad, Umer Asgher iD

Diagnostics · 12.11.2025 · 15 (22) · 2871

DOI: 10.3390/diagnostics15222871 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Diagnostics, online erschienen am 12.11.2025. Verfasst von Laraib Umer und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Mechanical and Manufacturing Engineering, National University of Sciences and Technology, Islamabad 44000, Pakistan. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 65 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Stress is a critical determinant of mental health, yet conventional monitoring approaches often rely on subjective self-reports or physiological signals that lack real-time responsiveness. Recent advances in large language models (LLMs) offer opportunities for speech-driven, adaptive stress detection, but existing systems are limited to retrospective text analysis, monolingual settings, or detection-only outputs.

Methods

We developed a real-time, speech-driven stress detection framework that integrates audio recording, speech-to-text conversion, and linguistic analysis using transformer-based LLMs. The system provides multimodal outputs, delivering recommendations in both text and synthesized speech. Nine LLM variants were evaluated on five benchmark datasets under zero-shot and few-shot learning conditions. Performance was assessed using accuracy, precision, recall, F1-score, and misclassification trends (false-negatives and false-positives). Real-time feasibility was analyzed through latency modeling, and user-centered validation was conducted across cross-domains.

Results

Few-shot fine-tuning improved model performance across all datasets, with Large Language Model Meta AI (LLaMA) and Robustly Optimized BERT Pretraining Approach (RoBERTa) achieving the highest F1-scores and reduced false-negatives, particularly for suicide risk detection. Latency analysis revealed a trade-off between responsiveness and accuracy, with delays ranging from ~2 s for smaller models to ~7.6 s for LLaMA-7B on 30 s audio inputs. Multilingual input support and multimodal output enhanced inclusivity. User feedback confirmed strong usability, accessibility, and adoption potential in real-world settings.

Conclusions

This study demonstrates that real-time, LLM-powered stress detection is both technically robust and practically feasible. By combining speech-based input, multimodal feedback, and user-centered validation, the framework advances beyond traditional detection only models toward scalable, inclusive, and deployment-ready digital mental health solutions.

Schlagwörter

Stichwörter der Autor:innen

clinical decision supportlarge language models (LLMs)personalized digital interventionsreal-time stress detectionspeech-based emotion recognitionstress classification

Autor:innen und Institutionen

  • School of Mechanical and Manufacturing Engineering, National University of Sciences and Technology, Islamabad 44000, Pakistan.
  • National Center of Artificial Intelligence, National University of Sciences and Technology, Islamabad 44000, Pakistan.
  • Department of Optometry and Orthoptics, Faculty of Rehabilitation and Allied Health Sciences, Riphah International University, Islamabad 44000, Pakistan.
  • School of Interdisciplinary Engineering and Sciences (SINES), National University of Sciences and Technology (NUST), Islamabad 44000, Pakistan.
  • Laboratory of Human Factors and Automation in Aviation, Department of Air Transport, Faculty of Transportation Sciences, Czech Technical University in Prague (CTU), 128 00 Prague, Czech Republic.

ORCID-Profile: Yasar Ayaz, Hassan Imam, Umer Asgher

Zitieren

APA

Umer, L., Iqbal, J., Ayaz, Y., Imam, H., Ahmad, A., & Asgher, U. (2025). StressSpeak: A Speech-Driven Framework for Real-Time Personalized Stress Detection and Adaptive Psychological Support. Diagnostics, 15(22), 2871. https://doi.org/10.3390/diagnostics15222871

Vancouver

Umer L, Iqbal J, Ayaz Y, Imam H, Ahmad A, Asgher U. StressSpeak: A Speech-Driven Framework for Real-Time Personalized Stress Detection and Adaptive Psychological Support. Diagnostics. 2025;15(22):2871. doi:10.3390/diagnostics15222871

BibTeX
@article{Umer2025,
  author = {Umer, Laraib and Iqbal, Javaid and Ayaz, Yasar and Imam, Hassan and Ahmad, Adil and Asgher, Umer},
  title = {StressSpeak: A Speech-Driven Framework for Real-Time Personalized Stress Detection and Adaptive Psychological Support},
  journal = {Diagnostics},
  year = {2025},
  volume = {15},
  number = {22},
  pages = {2871},
  doi = {10.3390/diagnostics15222871}
}
Kostenlos starten →