{"ID":2861331,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.08586","arxiv_id":"2510.08586","title":"Dynamic Stress Detection: A Study of Temporal Progression Modelling of Stress in Speech","abstract":"Detecting psychological stress from speech is critical in high-pressure settings. While prior work has leveraged acoustic features for stress detection, most treat stress as a static label. In this work, we model stress as a temporally evolving phenomenon influenced by historical emotional state. We propose a dynamic labelling strategy that derives fine-grained stress annotations from emotional labels and introduce cross-attention-based sequential models, a Unidirectional LSTM and a Transformer Encoder, to capture temporal stress progression. Our approach achieves notable accuracy gains on MuSE (+5%) and StressID (+18%) over existing baselines, and generalises well to a custom real-world dataset. These results highlight the value of modelling stress as a dynamic construct in speech.","short_abstract":"Detecting psychological stress from speech is critical in high-pressure settings. While prior work has leveraged acoustic features for stress detection, most treat stress as a static label. In this work, we model stress as a temporally evolving phenomenon influenced by historical emotional state. We propose a dynamic l...","url_abs":"https://arxiv.org/abs/2510.08586","url_pdf":"https://arxiv.org/pdf/2510.08586v2","authors":"[\"Vishakha Lall\",\"Yisi Liu\"]","published":"2025-10-02T06:30:44Z","proceeding":"eess.AS","tasks":"[\"eess.AS\",\"cs.AI\",\"cs.CL\",\"cs.SD\"]","methods":"[\"Transformer\"]","has_code":false}