Spectral Modification Based Data Augmentation For Improving End-to-End ASR For Children's Speech

Abstract

Training a robust Automatic Speech Recognition (ASR) system for children's speech recognition is a challenging task due to inherent differences in acoustic attributes of adult and child speech and scarcity of publicly available children's speech dataset. In this paper, a novel segmental spectrum warping and perturbations in formant energy are introduced, to generate a children-like speech spectrum from that of an adult's speech spectrum. Then, this modified adult spectrum is used as augmented data to improve end-to-end ASR systems for children's speech recognition. The proposed data augmentation methods give 6.5% and 6.1% relative reduction in WER on children dev and test sets respectively, compared to the vocal tract length perturbation (VTLP) baseline system trained on Librispeech 100 hours adult speech dataset. When children's speech data is added in training with Librispeech set, it gives a 3.7 % and 5.1% relative reduction in WER, compared to the VTLP baseline system.

A WAV2VEC2-Based Experimental Study on Self-Supervised Learning Methods to Improve Child Speech Recognition

“…Some of the other popular augmentation approaches include Vocal Tract Length Perturbation [21], Fundamental frequency feature normalization [22], out-ofdomain data augmentation using Stochastic Feature Mapping (SFM) [23], and data processing-based augmentations [24] such as Speed Perturbation, Pitch Perturbation, Tempo Perturbation, Volume Perturbation, Reverberation Perturbation, and Spectral Perturbation. Spectrogram Augmentation also seems promising for improving the performance of ASR systems [25], [26]. Each of these methods shows improvements in child ASR accuracy, however, they still require corresponding labeled annotations to speech data.…”