Acoustic Training from Heterogeneous Data Sources: Experiments in Mandarin Conversational Telephone Speech Transcription

Download: PDF.

“Acoustic Training from Heterogeneous Data Sources: Experiments in Mandarin Conversational Telephone Speech Transcription” by S. Tsakalidis and W. Byrne. In IEEE Conference on Acoustics, Speech and Signal Processing, 2005.

Abstract

In this paper we investigate the use of heterogeneous data sources for acoustic training. We describe an acoustic normalization procedure for enlarging an ASR acoustic training set with out-of-domain acoustic data. A larger in-domain training set is created by effectively transforming the out-of-domain data before incorporation in training. Baseline experimental results in Mandarin conversational telephone speech transcription show that a simple attempt to add out-of-domain data degrades performance. Preliminary experiments assess the effectiveness of the proposed cross-corpus acoustic normalization.

Download: PDF.

BibTeX entry:

@inproceedings{tsakalidis:icassp05,
   author = {S. Tsakalidis and W. Byrne},
   title = {Acoustic Training from Heterogeneous Data Sources: Experiments
	in {M}andarin Conversational Telephone Speech Transcription},
   booktitle = {IEEE Conference on Acoustics, Speech and Signal Processing},
   pages = {(4 pages)},
   year = {2005}
}

Back to Bill Byrne publications.