Fundamental Frequency Feature Normalization and Data Augmentation for Child Speech Recognition

TL;DR

This paper introduces novel fundamental frequency-based feature normalization and data augmentation techniques for child speech recognition, significantly improving accuracy by adapting adult speech models and addressing data scarcity issues.

Contribution

It presents a new frequency shift method in the Mel domain for feature normalization and data augmentation tailored for child ASR systems, enhancing performance.

Findings

19.3% relative WER reduction over baseline

Achieved best WER on OGI Kids' Speech Corpus

Effective adaptation of adult speech models for child speech

Abstract

Automatic speech recognition (ASR) systems for young children are needed due to the importance of age-appropriate educational technology. Because of the lack of publicly available young child speech data, feature extraction strategies such as feature normalization and data augmentation must be considered to successfully train child ASR systems. This study proposes a novel technique for child ASR using both feature normalization and data augmentation methods based on the relationship between formants and fundamental frequency ($f_o$). Both the $f_o$ feature normalization and data augmentation techniques are implemented as a frequency shift in the Mel domain. These techniques are evaluated on a child read speech ASR task. Child ASR systems are trained by adapting a BLSTM-based acoustic model trained on adult speech. Using both $f_o$ normalization and data augmentation results in a relative word error rate (WER) improvement of 19.3% over the baseline when tested on the OGI Kids' Speech Corpus, and the resulting child ASR system achieves the best WER currently reported on this corpus.