An Overview of Automatic Speaker Recognition in Adverse Acoustic Environment

Abstract

In the last 50 years, the task of automatic speaker recognition has undergone intense development and is still a popular topic. New ways are also being sought to implement speaker recognition tools in order to improve the educational process. These include support systems and services, such as recommendation and personalization-based content delivery, personalized speech assistant for human-computer interaction, or improved speech recognition systems based on speaker adaptation. Applications in the real environment stress upon technological issues such as dealing with acoustical environment variability, background noise, simultaneous speech or speaker's behaviour. The article provides a review of the state-of-the-art techniques for automatic speaker recognition with focus on its deployment in real and adverse acoustic conditions.