Algorithmic Exploration of American English Dialects

Aksënova, Alëna; Bruguier, Antoine; Ritchart-Scott, Amanda; Mendlovic, Uri

doi:10.1109/icassp40776.2020.9053751

Cited by 4 publications

(3 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Researchers also show that the quality of speech recognition might not be the same for various genders [8,9]. A study exploring the link between increased ASR errors and dialectal features demonstrated that geographic areas with increased WER might significantly overlap with regions that have certain dialectal features [4]. For example, higher WER scores are observed in the Southern United States, likely due to certain linguistic features of AAVE and Southern American English spoken in those regions.…”

Section: Benchmarking Asr Systems For Accentsmentioning

confidence: 99%

“…Literature reporting on results of various model evaluations shows that ASR models seem to frequently struggle with accented speech, be it achieving significantly higher word error rate (WER) for non-standard language varieties such as African American Vernacular English [2], code-switching [3], or observing increased WER in certain geographic areas that could be correlated with regional dialects [4]. In turn, this can make speakers of certain linguistic varieties feel excluded, and might prompt them to try "standardizing" and slowing down their speech [5].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Accented Speech Recognition: Benchmarking, Pre-training, and Diverse Data

Aksënova¹,

Chen²,

Chiu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Building inclusive speech recognition systems is a crucial step towards developing technologies that speakers of all language varieties can use. Therefore, ASR systems must work for everybody independently of the way they speak. To accomplish this goal, there should be available data sets representing language varieties, and also an understanding of model configuration that is the most helpful in achieving robust understanding of all types of speech. However, there are not enough data sets for accented speech, and for the ones that are already available, more training approaches need to be explored to improve the quality of accented speech recognition. In this paper, we discuss recent progress towards developing more inclusive ASR systems, namely, the importance of building new data sets representing linguistic diversity, and exploring novel training approaches to improve performance for all users. We address recent directions within benchmarking ASR systems for accented speech, measure the effects of wav2vec 2.0 pre-training on accented speech recognition, and highlight corpora relevant for diverse ASR evaluations.

show abstract

Section: Benchmarking Asr Systems For Accentsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Accented Speech Recognition: Benchmarking, Pre-training, and Diverse Data

Aksënova¹,

Chen²,

Chiu³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Such regional variants may involve regional phonology ('get' rhymes with 'vet' in the North, and with 'fit' in the South), and even significant lexical and syntactic differences ('going/planning to' can be expressed as 'fixin' to' in the South). Aksënova et al (2020) has shown how such regional variation can be explored, and how it can impact ASR performance. Ideally, then, as many regional variants as possible should be covered by the ideal benchmark for a given language.…”

Section: Defining Slicesmentioning

confidence: 99%

How Might We Create Better Benchmarks for Speech Recognition?

Aksënova¹,

Esch²,

Golik³

2021

Proceedings of the 1st Workshop on Benchmarking: Past, Present and Future

View full text Add to dashboard Cite

The applications of automatic speech recognition (ASR) systems are proliferating, in part due to recent significant quality improvements. However, as recent work indicates, even state-of-the-art speech recognition systems -some which deliver impressive benchmark results, struggle to generalize across use cases. We review relevant work, and, hoping to inform future benchmark development, outline a taxonomy of speech recognition use cases, proposed for the next generation of ASR benchmarks. We also survey work on metrics, in addition to the de facto standard Word Error Rate (WER) metric, and we introduce a versatile framework designed to describe interactions between linguistic variation and ASR performance metrics.

show abstract