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

Today, many state-of-the-art automatic speech recognition (ASR) systems applyall-neural models that map audio to word sequences trained end-to-end along oneglobal optimisation criterion in a fully data driven fashion. These modelsallow high precision ASR for domains and words represented in the trainingmaterial but have difficulties recognising words that are rarely or not at allrepresented during training, i.e. trending words and new named entities. Inthis paper, we use a text-to-speech (TTS) engine to provide synthetic audio forout-of-vocabulary (OOV) words. We aim to boost the recognition accuracy of arecurrent neural network transducer (RNN-T) on OOV words by using the extraaudio-text pairs, while maintaining the performance on the non-OOV words.Different regularisation techniques are explored and the best performance isachieved by fine-tuning the RNN-T on both original training data and extrasynthetic data with elastic weight consolidation (EWC) applied on the encoder.This yields a 57% relative word error rate (WER) reduction on utterancescontaining OOV words without any degradation on the whole test set.

Using Synthetic Audio to Improve The Recognition of Out-Of-Vocabulary Words in End-To-End ASR Systems