基于共振峰位置加权光谱特征的情感识别外文翻译.docxVIP

外文原文 Formant position based weighted spectral features for emotion recognition Abstract In this paper, we propose novel spectrally weighted mel-frequency cepstral coefficient (WMFCC) features for emotion recognition from speech. The idea is based on the fact that formant locations carry emotion-related information, and therefore critical spectral bands around formant locations can be emphasized during the calculation of MFCC features. The spectral weighting is derived from the normalized inverse harmonic mean function of the line spectral frequency (LSF) features, which are known to be localized around formant frequencies. The above approach can be considered as an early data fusion of spectral content and formant location information. We also investigate methods for late decision fusion of unimodal classifiers. We evaluate the proposed WMFCC features together with the standard spectral and prosody features using HMM based classifiers on the spontaneous FAU Aibo emotional speech corpus. The results show that unimodal classifiers with the WMFCC features perform significantly better than the classifiers with standard spectral features. Late decision fusion of classifiers provide further significant performance improvements. Keywords: Emotion recognition; Emotional speech classification; Spectral features; Formant frequency; Line spectral frequency; Decision fusion. Introduction Emotion-sensitive machine intelligence is a basic requirement for more natural human–computer interaction. In this sense, the orientation of research on emotional speech processing shifts from the analysis of acted towards spontaneous speech for advanced real-life applications in human–machine interaction systems (Ververidis and Kotropoulos, 2006; Batliner et al., 2003). The wide use of telecommunication services and multimedia devices will require human-centered designs instead of computer centered ones. Consequently, accurate perception of the user’s affective state by computer systems will b