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

Contextual knowledge is essential for reducing speech recognition errors onhigh-valued long-tail words. This paper proposes a novel tree-constrainedpointer generator (TCPGen) component that enables end-to-end ASR models to biastowards a list of long-tail words obtained using external contextualinformation. With only a small overhead in memory use and computation cost,TCPGen can structure thousands of biasing words efficiently into a symbolicprefix-tree and creates a neural shortcut between the tree and the final ASRoutput to facilitate the recognition of the biasing words. To enhance TCPGen,we further propose a novel minimum biasing word error (MBWE) loss that directlyoptimises biasing word errors during training, along with a biasing-word-drivenlanguage model discounting (BLMD) method during the test. All contextual ASRsystems were evaluated on the public Librispeech audiobook corpus and the datafrom the dialogue state tracking challenges (DSTC) with the biasing listsextracted from the dialogue-system ontology. Consistent word error rate (WER)reductions were achieved with TCPGen, which were particularly significant onthe biasing words with around 40\% relative reductions in the recognition errorrates. MBWE and BLMD further improved the effectiveness of TCPGen and achievedmore significant WER reductions on the biasing words. TCPGen also achievedzero-shot learning of words not in the audio training set with large WERreductions on the out-of-vocabulary words in the biasing list.

Minimising Biasing Word Errors for Contextual ASR with the Tree-Constrained Pointer Generator