Computational Pronunciation Analysis in Sung Utterances
Information Retrieval
2021-06-22 v1 Sound
Audio and Speech Processing
Abstract
Recent automatic lyrics transcription (ALT) approaches focus on building stronger acoustic models or in-domain language models, while the pronunciation aspect is seldom touched upon. This paper applies a novel computational analysis on the pronunciation variances in sung utterances and further proposes a new pronunciation model adapted for singing. The singing-adapted model is tested on multiple public datasets via word recognition experiments. It performs better than the standard speech dictionary in all settings reporting the best results on ALT in a capella recordings using n-gram language models. For reproducibility, we share the sentence-level annotations used in testing, providing a new benchmark evaluation set for ALT.
Cite
@article{arxiv.2106.10977,
title = {Computational Pronunciation Analysis in Sung Utterances},
author = {Emir Demirel and Sven Ahlback and Simon Dixon},
journal= {arXiv preprint arXiv:2106.10977},
year = {2021}
}