Combining regression and classification methods for improving automatic speaker age recognition

We present a novel approach to automatic speaker age classification, which combines regression and classification to achieve competitive classification accuracy on telephone speech. Support vector machine regression is used to generate finer age estimates, which are combined with the posterior probabilities of well-trained discriminative gender classifiers to predict both the age and gender of a speaker. We show that this combination performs better than direct 7-class classifiers. The regressors and classifiers are trained using longterm features such as pitch and formants, as well as short-term (frame-based) features derived from MAP adaptation of GMMs that were trained on MFCCs.

1. INTRODUCTION

Speech is the most natural means of communication among humans. We not only share semantic concepts through our speech but also convey so-called paralinguistic information, such as the identity of the speaker, the gender, the approximate age, emotional state etc. This information guides the way in which we communicate with each other and is also considered to be important for human-computer communications, such as in interactive voice response (IVR) sys-tems. Angry customers can, for example, be detected and approached differently by connecting them to a sympathetic human operator. In much the same way, dialogues, as well as background music or advertisements played while waiting for an operator, can be adapted based on other paralinguistic cues such as the speaker's age or gender [1].