Emotion Recognition in Speech Using Neural Networks

Abstract

Emotion recognition in speech is a topic on which little research has been done to-date. In this paper, we discuss why emotion recognition in speech is a significant and applicable research topic, and present a system for emotion recognition using oneclass-in-one neural networks. By using a large database of phoneme balanced words, our system is speaker-and context-independent. We achieve a recognition rate of approximately 50% when testing eight emotions.

A user-independent real-time emotion recognition system for software agents in domestic environments

“…Hitherto, the identification and classification of emotional changes has achieved mixed results ranging from 60-95.5% detection accuracy for facial recognition (Avent et al, 1994;Rosenblum et al, 1994;Sun et al, 2004;Bartlett et al, 2003;Anderson and McOwan, 2003;De Silva and Hui, 2003) to 50-87.5% for speech recognition (Nicholson et al, 2000;Tsuyoshi and Shinji, 1999), and 72% in bimodal recognition (face and speech) (De Silva and Ng, 1999). In physiological emotion detection some of the best results have been achieved by Kim et al (2002) (Avent et al, 1994;Rosenblum et al, 1994;Sun et al, 2004;Bartlett et al, 2003;Anderson and McOwan, 2003;De Silva and Hui, 2003;Nicholson et al, 2000) and advanced statistical mechanisms (Tsuyoshi and Shinji, 1999;De Silva and Ng, 1999;Kim et al, 2002;Nasoz et al, 2003;Picard et al, 2001).…”

“…In physiological emotion detection some of the best results have been achieved by Kim et al (2002) (Avent et al, 1994;Rosenblum et al, 1994;Sun et al, 2004;Bartlett et al, 2003;Anderson and McOwan, 2003;De Silva and Hui, 2003;Nicholson et al, 2000) and advanced statistical mechanisms (Tsuyoshi and Shinji, 1999;De Silva and Ng, 1999;Kim et al, 2002;Nasoz et al, 2003;Picard et al, 2001).…”

Evaluation of influence of spectral and prosodic features on GMM classification of Czech and Slovak emotional speech

“…Included in this category is also a separate group of features constituting the voice quality parameters: jitter, shimmer [11], Hammarberg index [12], LiljencrantsFant features [13], and spectral tilt [14]. All mentioned speech identification systems and classifiers are usually based on statistical approach, using the discriminative or artificial neural networks [15,16], hidden Markov models (HMM) [17], or Gaussian mixture models (GMM) [18,19]. Spectral features like MFCC together with energy and prosodic parameters are most commonly used in GMM emotional speech classification [20].…”

Automatic Recognition of Emotional States From Human Speeches

“…As there is a lack of precise definition and models for emotions, automatic recognition of emotions has been a challenging task to researchers. Indeed, research on speech based emotion recognition has been undertaken by many for around two decades (Amir, 2001;Clavel et al, 2004;Cowie & Douglas-Cowie, 1996;Cowie et al, 2001;Dellaert et al, 1996;Lee & Narayanan, 2005;Morrison et al, 2007;Nguyen & Bass, 2005;Nicholson et al, 1999;Petrushin, 1999;Petrushin, 2000;Scherer, 2000;Ser et al, 2008;Ververidis & Kotropoulos, 2006;Yu et al, 2001;Zhou et al, 2006). In engineering, speech emotion recognition has been formulated as a pattern recognition problem that involves feature extraction and emotion classification.…”