A Very Low Resource Language Speech Corpus for Computational Language Documentation Experiments

Godard, P.; Adda, G.; Adda-Decker, M.; Benjumea, J.; Besacier, L.; Cooper-Leavitt, J.; Kouarata, G-N.; Lamel, L.; Maynard, H.; Mueller, M.; Rialland, A.; Stueker, S.; Yvon, F.; Zanon-Boito, M.

doi:10.48550/arxiv.1710.03501

Cited by 10 publications

(14 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The same datasets of [9] will be used: a corpus in Kunwinjku which consists of 301 utterances aligned with an orthographic transcription and a forced alignment created using the MAUS forced aligner [20], and a corpus in Mboshi which consists of 5130 utterances elicited from text, with orthographic transcription and a forced alignment at the word level [21].…”

Section: Datamentioning

confidence: 99%

Spoken Term Detection Methods for Sparse Transcription in Very Low-resource Settings

Ferrand¹,

Bird²,

Besacier³

2021

Preprint

View full text Add to dashboard Cite

We investigate the efficiency of two very different spoken term detection approaches for transcription when the available data is insufficient to train a robust ASR system. This work is grounded in very low-resource language documentation scenario where only few minutes of recording have been transcribed for a given language so far. Experiments on two oral languages show that a pretrained universal phone recognizer, fine-tuned with only a few minutes of target language speech, can be used for spoken term detection with a better overall performance than a dynamic time warping approach. In addition, we show that representing phoneme recognition ambiguity in a graph structure can further boost the recall while maintaining high precision in the low resource spoken term detection task.

show abstract

Section: Datamentioning

confidence: 99%

Spoken Term Detection Methods for Sparse Transcription in Very Low-resource Settings

Ferrand¹,

Bird²,

Besacier³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In this section, we evaluate generalization of multilingual ST models by performing transfer learning to a very low-resource ST task. We used Mboshi-French corpus 14 [22], which contains 4.4-hours of spoken utterances and the corresponding Mboshi transcriptions and French translations. Mboshi [49] is a Bantu C25 language spoken in Congo-Brazzaville and does not have standard orthography.…”

Section: Pre-training With the Asr Encodermentioning

confidence: 99%

“…We evaluate one-to-many (O2M) and many-to-many (M2M) translations by combining these corpora and confirm significant improvements by multilingual training in both scenarios. Next, we evaluate the generalization of multilingual E2E-ST models by performing transfer learning to a very low-resource ST task: Mboshi (Bantu C25)→Fr corpus (4.4 hours) [22]. We show that multilingual pretraining of the seed E2E-ST models improves the performance in the low-resource language pair unseen during training, compared to bilingual pre-training.…”

Section: Introductionmentioning

confidence: 98%

Multilingual End-to-End Speech Translation

Inaguma¹,

Duh²,

Kawahara³

et al. 2019

Preprint

View full text Add to dashboard Cite

In this paper, we propose a simple yet effective framework for multilingual end-to-end speech translation (ST), in which speech utterances in source languages are directly translated to the desired target languages with a universal sequence-to-sequence architecture. While multilingual models have shown to be useful for automatic speech recognition (ASR) and machine translation (MT), this is the first time they are applied to the end-to-end ST problem. We show the effectiveness of multilingual end-to-end ST in two scenarios: one-to-many and many-to-many translations with publicly available data. We experimentally confirm that multilingual end-to-end ST models significantly outperform bilingual ones in both scenarios. The generalization of multilingual training is also evaluated in a transfer learning scenario to a very low-resource language pair. All of our codes and the database are publicly available to encourage further research in this emergent multilingual ST topic 1 .

show abstract

“…Mboshi-French: Mboshi (Bantu C25 in the Guthrie classification) is a language spoken in Congo-Brazzaville, without standard orthography. We use a corpus of 5517 parallel utterances (about 4.4 hours of audio) collected from three native speakers using the LIG-Aikuma app for the BULB project [5,7]. The corpus provides non-standard grapheme transcriptions (produced by linguists to be close to the language phonology) as well as French translations.…”

Section: Datamentioning

confidence: 99%

“…One is Ainu, a severely endangered language, with translations in English. We also experiment on a recently collected speech corpus of Mboshi [7], with translations in French.…”

Section: Introductionmentioning

confidence: 99%

Leveraging Translations for Speech Transcription in Low-resource Settings

Anastasopoulos¹,

Chiang

2018

Interspeech 2018

View full text Add to dashboard Cite

Recently proposed data collection frameworks for endangered language documentation aim not only to collect speech in the language of interest, but also to collect translations into a highresource language that will render the collected resource interpretable. We focus on this scenario and explore whether we can improve transcription quality under these extremely lowresource settings with the assistance of text translations. We present a neural multi-source model and evaluate several variations of it on three low-resource datasets. We find that our multi-source model with shared attention outperforms the baselines, reducing transcription character error rate by up to 12.3%.

show abstract

A Very Low Resource Language Speech Corpus for Computational Language Documentation Experiments

Cited by 10 publications

References 14 publications

Spoken Term Detection Methods for Sparse Transcription in Very Low-resource Settings

Spoken Term Detection Methods for Sparse Transcription in Very Low-resource Settings

Multilingual End-to-End Speech Translation

Leveraging Translations for Speech Transcription in Low-resource Settings

Contact Info

Product

Resources

About