Design and Implementation of Smartphone Applications for Speaker Count and Gender Recognition
The technological evolution of smartphones, combined with their increasing diffusion, gives mobile network providers the opportunity to provide real-time, context-aware services. The number of speakers in the surroundings of the smartphone can be useful context information. In this paper, a speaker count method based on pitch estimation designed to recognize single-speaker samples from two-speaker samples is presented. In addition, a method for speaker gender recognition is described as well. An implementation of both methods for Symbian OS smartphones is also presented.
KeywordsFeature Vector Probability Distribution Function Gaussian Mixture Model Female Speaker Audio Sample
- 2.Gerhard D (2003) Pitch extraction and fundamental frequency: history and current techniques. Ph.D. Dissertation, Department of Computer Science, University of ReginaGoogle Scholar
- 3.Web site: http://www.phon.ucl.ac.uk/courses/spsci/matlab/lect10.html. (2009) University College of London, Department of Phonetics and Linguistics