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

While end-to-end models have shown great success on the Automatic SpeechRecognition task, performance degrades severely when target sentences arelong-form. The previous proposed methods, (partial) overlapping inference areshown to be effective on long-form decoding. For both methods, word error rate(WER) decreases monotonically when overlapping percentage decreases. Settingaside computational cost, the setup with 50% overlapping during inference canachieve the best performance. However, a lower overlapping percentage has anadvantage of fast inference speed. In this paper, we first conductcomprehensive experiments comparing overlapping inference and partialoverlapping inference with various configurations. We then proposeVoice-Activity-Detection Overlapping Inference to provide a trade-off betweenWER and computation cost. Results show that the proposed method can achieve a20% relative computation cost reduction on Librispeech and Microsoft SpeechLanguage Translation long-form corpus while maintaining the WER performancewhen comparing to the best performing overlapping inference algorithm. We alsopropose Soft-Match to compensate for similar words mis-aligned problem.

VADOI:Voice-Activity-Detection Overlapping Inference For End-to-end Long-form Speech Recognition