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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent advances with self-supervised learning have allowed speech recognitionsystems to achieve state-of-the-art (SOTA) word error rates (WER) whilerequiring only a fraction of the labeled training data needed by itspredecessors. Notwithstanding, while such models achieve SOTA performance inmatched train/test conditions, their performance degrades substantially whentested in unseen conditions. To overcome this problem, strategies such as dataaugmentation and/or domain shift training have been explored. Available models,however, are still too large to be considered for edge speech applications onresource-constrained devices, thus model compression tools are needed. In thispaper, we explore the effects that train/test mismatch conditions have onspeech recognition accuracy based on compressed self-supervised speech models.In particular, we report on the effects that parameter quantization and modelpruning have on speech recognition accuracy based on the so-called robustwav2vec 2.0 model under noisy, reverberant, and noise-plus-reverberationconditions.

On the Impact of Quantization and Pruning of Self-Supervised Speech Models for Downstream Speech Recognition Tasks "In-the-Wild''