Enhanced speaker recognition based on score level fusion of AHS and HMM

Abstract

Speaker recognition history dates back to some four decades, and yet it has not been reliable enough to be considered as a standalone security system. This paper focuses on the enhancement of speaker recognition through fusion of likelihood scores generated by arithmetic harmonic sphericity (AHS) and hidden Markov model (HMM) techniques. Due to the contrastive nature of AHS and HMM, we have observed a significant performance improvement of 22% and 6% true acceptance rate at 5% false acceptance rate, when this fusion technique was evaluated on two different datasets - YOHO and USF multimodal biometric dataset, respectively. Performance enhancement has been achieved on both the datasets, however performance on YOHO was comparatively higher than that on USF dataset, owing to the fact that USF dataset is a noisy outdoor dataset whereas YOHO is an indoor dataset.