基於生理訊號變化即時偵測音樂誘發情緒研究

Abstract

音樂在日常生活扮演著重要的角色，一首扣人心弦的音樂，會使聽者產生共鳴，激起聽者情緒的波動，不同的音樂會促使聽者出現不同的情緒反應，本實驗室歷年來研究的內容趨勢皆以分析音樂情緒為主軸，辨識每首音樂本身所代表的情緒，然而對於人們聽音樂的真實情緒是否與系統辨識出的音樂情緒一樣仍存有不確定性，因此，本研究主要銜接過去實驗室的音樂情緒研究，利用生理感測器測量測試者聽音樂前與聽音樂後時的生理反應，將生理感測器測得的生理參數擷取加以分析，目的為觀察測試者聆聽音樂過程中，其情緒之生理反應表現，是否與數理方式分析出的音樂情緒結果相符，並比較聆聽音前樂與聽音樂時的生理訊號有何相異。於數種生理訊號中，本研究將採用心率值(Heart Rate)與心率變異度(Heart rate variability)以及肌電圖訊號(Electromyography)三種生理參數，來進行音樂與人類情緒變化之間的關係研究。

Emotional responses varied greatly according to musical genre and type of response. In recent years, our laboratory researches are focused on mathematical analysis of music emotion to recognize the mood variation in music listening. However, there are not sure the music’s emotion which recognized by mathematical analysis are the same as feelings of humans. This research studies musically induced emotions as an experiential phenomenon. Emotions identified with physiological manifestations, without excluding those emotional states that may not have overt expressions but still represent highly characteristic reactions to music. Three kind of physiological signals: heart rate(HR), heart rate variability(HRV), and electromyography(EMG) were used to observe the relationship between emotions evoked by the sound of music and physiological responses. The study measured electromyography, heart rate, and heart rate variability in real time when participants listening to music, and convert the physiology signals into power spectrum to observe relationship between mathematical analysis of music emotion and physiology signals of humans emotion. Also, many physiological responses to music other than emotional reaction are possible, and not all individuals react emotionally to music, we believe that a systematic investigation of musically induced emotions may benefit from a focus on feeling.