Proceedings of the First Workshop on Natural Language Processing for Indigenous Languages of the Americas 2021
DOI: 10.18653/v1/2021.americasnlp-1.20
|View full text |Cite
|
Sign up to set email alerts
|

Explicit Tone Transcription Improves ASR Performance in Extremely Low-Resource Languages: A Case Study in Bribri

Abstract: Linguistic tone is transcribed for input into ASR systems in numerous ways. This paper shows a systematic test of several transcription styles, using as an example the Chibchan language Bribri, an extremely low-resource language from Costa Rica. The most successful models separate the tone from the vowel, so that the ASR algorithms learn tone patterns independently. These models showed improvements ranging from 4% to 25% in character error rate (CER), and between 3% and 23% in word error rate (WER). This is tr… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
1
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
2
1

Relationship

0
3

Authors

Journals

citations
Cited by 3 publications
(1 citation statement)
references
References 38 publications
0
1
0
Order By: Relevance
“…But for Indigenous and other under-resourced languages, creating ASR entails first collecting a large mass of already-transcribed data -in addition to the data for the sociolinguistic experiment -so that a new ASR model can be trained. This process is still difficult and expensive, requiring (i) linguists and community experts to settle on an orthographic or phonetic representation of the language, (ii) human experts to transcribe hours of recordings (from 4 to 100), and (iii) programmers to train the system on specialised computer servers (Adams et al, 2019;Besacier et al, 2014;Coto-Solano, 2021;Foley et al, 2018;Gupta & Boulianne, 2020;Levow et al, 2021;Matsuura et al, 2020;Partanen et al, 2020;Prud'hommeaux et al, 2021;Zahrer et al, 2020;Zevallos et al, 2020).…”
Section: Automatic Speech Recognition For Sociophoneticsmentioning
confidence: 99%
“…But for Indigenous and other under-resourced languages, creating ASR entails first collecting a large mass of already-transcribed data -in addition to the data for the sociolinguistic experiment -so that a new ASR model can be trained. This process is still difficult and expensive, requiring (i) linguists and community experts to settle on an orthographic or phonetic representation of the language, (ii) human experts to transcribe hours of recordings (from 4 to 100), and (iii) programmers to train the system on specialised computer servers (Adams et al, 2019;Besacier et al, 2014;Coto-Solano, 2021;Foley et al, 2018;Gupta & Boulianne, 2020;Levow et al, 2021;Matsuura et al, 2020;Partanen et al, 2020;Prud'hommeaux et al, 2021;Zahrer et al, 2020;Zevallos et al, 2020).…”
Section: Automatic Speech Recognition For Sociophoneticsmentioning
confidence: 99%