Hidden Markov model-based speech emotion recognition

Abstract

In this contribution we introduce speech emotion recognition by use of continuous hidden Markov models. Two methods are propagated and compared throughout the paper. Within the first method a global statistics framework of an utterance is classified by Gaussian mixture models using derived features of the raw pitch and energy contour of the speech signal. A second method introduces increased temporal complexity applying continuous hidden Markov models considering several states using low-level instantaneous features instead of global statistics. The paper addresses the design of working recognition engines and results achieved with respect to the alluded alternatives. A speech corpus consisting of acted and spontaneous emotion samples in German and English language is described in detail. Both engines have been tested and trained using this equivalent speech corpus. Results in recognition of seven discrete emotions exceeded 86% recognition rate. As a basis of comparison the similar judgment of human deciders classifying the same corpus at 79.8% recognition rate was analyzed.

Development of Acoustic Model for Croatian Language Using HTK

“…Response of the subject is measured using different modalities, like psychophysiological responses (heart rate, blood pressure, skin conductance and temperature), but also through verbal and nonverbal responses. The analysis of vocal features can give important information about the emotional state of a person [11]. Emotions can be expressed non-verbally, through prosodic structure of an utterance, but also verbally, by directly expressing thoughts and feelings.…”

An evaluation of the robustness of existing supervised machine learning approaches to the classification of emotions in speech

“…Although utterance level approaches are the most common (Schuller et al, 2005;Cichosz and Slot, 2005;Oudeyer, 2003), segment based approaches are becoming more popular. Segment based approaches try to model the shape of acoustic contours more closely as in (Katz et al, 1996;Schuller et al, 2003;Batliner et al, 2003;Batliner et al, 2005;Rotaru and Litman, 2005). In all of the mentioned studies, a single speech corpus is used for training and testing a machine learned classifier.…”

Time-Scale Feature Extractions for Emotional Speech Characterization

“…This approach employs machine learning techniques such as Hidden Markov Models [13]. Speech and speaker recognition techniques: short-term features and statistical modeling (GMM, HMM) have been successfully combined with a traditional turn based level approach [15].…”

“…Indeed, the standard unit is the speaker turn level [12][13][14] which consists in the characterization of a whole sentence by a large number of features. This approach assumes that the emotional state is not changing during the speaker turn level.…”