A Study of Enhancement, Augmentation, and Autoencoder Methods for Domain Adaptation in Distant Speech Recognition
This work addresses the domain adaptation challenge in speech recognition for improving distant speech applications, but it is incremental as it reviews and compares existing methods rather than introducing new ones.
The paper tackles the problem of adapting close-talking speech recognizers to distant speech, where word error rates can degrade by up to 40% absolute, by reviewing and comparing methods like speech enhancement and autoencoders on the AMI dataset to quantify performance gaps and set a basis for future adaptation studies.
Speech recognizers trained on close-talking speech do not generalize to distant speech and the word error rate degradation can be as large as 40% absolute. Most studies focus on tackling distant speech recognition as a separate problem, leaving little effort to adapting close-talking speech recognizers to distant speech. In this work, we review several approaches from a domain adaptation perspective. These approaches, including speech enhancement, multi-condition training, data augmentation, and autoencoders, all involve a transformation of the data between domains. We conduct experiments on the AMI data set, where these approaches can be realized under the same controlled setting. These approaches lead to different amounts of improvement under their respective assumptions. The purpose of this paper is to quantify and characterize the performance gap between the two domains, setting up the basis for studying adaptation of speech recognizers from close-talking speech to distant speech. Our results also have implications for improving distant speech recognition.