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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

As for other forms of AI, speech recognition has recently been examined withrespect to performance disparities across different user cohorts. One approachto achieve fairness in speech recognition is to (1) identify speaker cohortsthat suffer from subpar performance and (2) apply fairness mitigation measurestargeting the cohorts discovered. In this paper, we report on initial findingswith both discovery and mitigation of performance disparities using data from aproduct-scale AI assistant speech recognition system. We compare cohortdiscovery based on geographic and demographic information to a more scalablemethod that groups speakers without human labels, using speaker embeddingtechnology. For fairness mitigation, we find that oversampling ofunderrepresented cohorts, as well as modeling speaker cohort membership byadditional input variables, reduces the gap between top- and bottom-performingcohorts, without deteriorating overall recognition accuracy.

Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparities