Fast transcription of speech in low-resource languages

2019/09/16 by Hasegawa-Johnson, Mark, Goudeseune, Camille, Levow, Gina-Anne
#68T10 #Computation and Language (cs.CL) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.1909.07285

Abstract

We present software that, in only a few hours, transcribes forty hours of recorded speech in a surprise language, using only a few tens of megabytes of noisy text in that language, and a zero-resource grapheme to phoneme (G2P) table. A pretrained acoustic model maps acoustic features to phonemes; a reversed G2P maps these to graphemes; then a language model maps these to a most-likely grapheme sequence, i.e., a transcription. This software has worked successfully with corpora in Arabic, Assam, Kinyarwanda, Russian, Sinhalese, Swahili, Tagalog, and Tamil.

Related

• Cost Analysis of Human-corrected Transcription for Predominately Oral Languages
• Enabling Interactive Transcription in an Indigenous Community
• Leveraging translations for speech transcription in low-resource settings
• A Very Low Resource Language Speech Corpus for Computational Language Documentation Experiments
• Spoken Term Detection Methods for Sparse Transcription in Very Low-resource Settings