Studies on noise robust automatic speech recognition

摘要

Noise in everyday acoustic environments such as cars, traffic environments, and cafeterias remains one of the main challenges in automatic speech recognition (ASR). As a research theme, it has received wide attention in conferences and scientific journals focused on speech technology. This article collection reviews both the classic and novel approaches suggested for noise robust ASR. The articles are literature reviews written for the spring 2009 seminar course on noise robust automatic speech recognition (course code T-61.6060) held at TKK.