Robust Automatic Speech Recognition in Real Life Environment using Active Noise Contro

Abstract

This paper presents the implementation of a robust automated speech recognition system when operated in real life environment by using active noise control. Filtered-x Least Mean Square Algorithm is used as a primary tool for performing Active Noise Control. Previous researches carried out on ANC using FxLMS were having distinct applications. Whereas, this research evaluates the performance of FxLMS algorithm specifically for the application of Automated Speech Recognition. Experiments were conducted and it was found that the performance of the ASR system showed considerable improvements in the presence of the ANC, when tested in noisy environments.