Scholarlink

<p>We propose a continuous stress-detection pipeline that integrates wavelet-based heart-rate-variability (HRV) features with behavioural signals in a staged ensemble. The method computes time-resolved HRV spectral indices (VLF/LF/HF) using wavelet decomposition and incorporates them with behavioural descriptors before training a Random-Forest-based ensemble on the fused feature set. Post-hoc SHAP analysis yields both dataset-level importance rankings and instance-level attributions, enabling clinicians and end users to interpret the rationale underlying each prediction. Evaluated on the WESAD dataset (<i>N</i>=15; ECG sampled at 700&#xa0;Hz; session duration ≈35–40&#xa0;min), the system achieved 83.0% accuracy on a held-out test split. SHAP consistently highlighted LF and HF approximate power as dominant predictors, while the wavelet plots revealed short-term spectral fluctuations aligned with well-known stress episodes, reinforcing the physiological validity of the reported findings.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An explainable multimodal deep learning approach for stress detection in emotion-aware systems

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月21日

查看全文
  • Sai H. Prajwal,
  • Sofia Singh,
  • Dipti Theng

摘要

We propose a continuous stress-detection pipeline that integrates wavelet-based heart-rate-variability (HRV) features with behavioural signals in a staged ensemble. The method computes time-resolved HRV spectral indices (VLF/LF/HF) using wavelet decomposition and incorporates them with behavioural descriptors before training a Random-Forest-based ensemble on the fused feature set. Post-hoc SHAP analysis yields both dataset-level importance rankings and instance-level attributions, enabling clinicians and end users to interpret the rationale underlying each prediction. Evaluated on the WESAD dataset (N=15; ECG sampled at 700 Hz; session duration ≈35–40 min), the system achieved 83.0% accuracy on a held-out test split. SHAP consistently highlighted LF and HF approximate power as dominant predictors, while the wavelet plots revealed short-term spectral fluctuations aligned with well-known stress episodes, reinforcing the physiological validity of the reported findings.