Earticle

초록

영어

An analysis of intonation contours extracted from Korean emotional speech using Momel is conducted. For the analysis, 360 sentence tokens of six emotions (happiness, anger, fear, sadness, neutral and boredom) produced by 6 actors (3 men and 3 women) are processed to extract acoustical feature values. The feature values include micro level features such as Pt (pitch point movement time), Pr (pitch point pitch range), Ps (pitch point pitch slope) and Pd (pitch point distance). In addition, sentence level features are included, such as Sr (sentence pitch range), Sf (sentence final pitch range) and Syl (syllables per second). Some micro level features (Pt and Pd) and all sentence level features (Sr, Ss, Sf, and Syl) show to be affected by emotions. Gender differences are not found in any of the micro level features whereas all of the sentence level features are affected by gender. In the same vein as the previous study on German emotional speech (Yi, 2014a), this study on Korean emotional speech is also showing that the emotions with high level arousal are differentiated from those with low level arousal by the acoustical features.

목차

Abstract
 1. 서론
 2. 실험
  2.1. 감정 음성 데이터베이스
  2.2. 연구 방법
  2.3. 결과 및 분석
 3. 결론
 참고문헌

저자

• 이서배 [ So Pae Yi | Humphreys West Elementary School, Department of Defense Education Activity ]

참고문헌