Making More of Little Data: Improving Low-Resource Automatic Speech Recognition Using Data Augmentation

1 code implementation • 18 May 2023 • Martijn Bartelds, Nay San, Bradley McDonnell, Dan Jurafsky, Martijn Wieling

For Gronings, for which there was a pre-existing text-to-speech (TTS) system available, we also examined the use of TTS to generate ASR training data from text-only sources.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +2

Leveraging supplementary text data to kick-start automatic speech recognition system development with limited transcriptions

no code implementations • 9 Feb 2023 • Nay San, Martijn Bartelds, Blaine Billings, Ella de Falco, Hendi Feriza, Johan Safri, Wawan Sahrozi, Ben Foley, Bradley McDonnell, Dan Jurafsky

We perform experiments using 10 minutes of transcribed speech from English (for replicating prior work) and two additional pairs of languages differing in the availability of supplemental text data: Gronings and Frisian (~7. 5M token corpora available), and Besemah and Nasal (only small lexica available).

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

Quantifying Language Variation Acoustically with Few Resources

1 code implementation • NAACL 2022 • Martijn Bartelds, Martijn Wieling

Deep acoustic models represent linguistic information based on massive amounts of data.

Dynamic Time Warping

Automated speech tools for helping communities process restricted-access corpora for language revival efforts

no code implementations • ComputEL (ACL) 2022 • Nay San, Martijn Bartelds, Tolúlopé Ògúnrèmí, Alison Mount, Ruben Thompson, Michael Higgins, Roy Barker, Jane Simpson, Dan Jurafsky

An even narrower bottleneck occurs for recordings with access constraints, such as language that must be vetted or filtered by authorised community members before annotation can begin.

Action Detection Activity Detection +6

Adapting Monolingual Models: Data can be Scarce when Language Similarity is High

1 code implementation • Findings (ACL) 2021 • Wietse de Vries, Martijn Bartelds, Malvina Nissim, Martijn Wieling

For many (minority) languages, the resources needed to train large models are not available.

POS POS Tagging +1

Leveraging pre-trained representations to improve access to untranscribed speech from endangered languages

1 code implementation • 26 Mar 2021 • Nay San, Martijn Bartelds, Mitchell Browne, Lily Clifford, Fiona Gibson, John Mansfield, David Nash, Jane Simpson, Myfany Turpin, Maria Vollmer, Sasha Wilmoth, Dan Jurafsky

Surprisingly, the English model outperformed the multilingual model on 4 Australian language datasets, raising questions around how to optimally leverage self-supervised speech representations for QbE-STD.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

Neural Representations for Modeling Variation in Speech

1 code implementation • 25 Nov 2020 • Martijn Bartelds, Wietse de Vries, Faraz Sanal, Caitlin Richter, Mark Liberman, Martijn Wieling

We show that speech representations extracted from a specific type of neural model (i. e. Transformers) lead to a better match with human perception than two earlier approaches on the basis of phonetic transcriptions and MFCC-based acoustic features.