English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Self-supervised pre-trained speech models have strongly improved speechrecognition, yet they are still sensitive to domain shifts and accented oratypical speech. Many of these models rely on quantisation or clustering tolearn discrete acoustic units. We propose to correct the discovered discreteunits for accented speech back to a standard pronunciation in an unsupervisedmanner. A masked language model is trained on discrete units from a standardaccent and iteratively corrects an accented token sequence by maskingunexpected cluster sequences and predicting their common variant. Small accentadapter blocks are inserted in the pre-trained model and fine-tuned bypredicting the corrected clusters, which leads to an increased robustness ofthe pre-trained model towards a target accent, and this without supervision. Weare able to improve a state-of-the-art HuBERT Large model on a downstreamaccented speech recognition task by altering the training regime with theproposed method.

Unsupervised Accent Adaptation Through Masked Language Model Correction Of Discrete Self-Supervised Speech Units