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

The integration of Language Models (LMs) has proven to be an effective way toaddress domain shifts in speech recognition. However, these approaches usuallyrequire a significant amount of target domain text data for the training ofLMs. Different from these methods, in this work, with only a domain-specifictext prompt, we propose two zero-shot ASR domain adaptation methods usingLLaMA, a 7-billion-parameter large language model (LLM). LLM is used in twoways: 1) second-pass rescoring: reranking N-best hypotheses of a given ASRsystem with LLaMA; 2) deep LLM-fusion: incorporating LLM into the decoder of anencoder-decoder based ASR system. Experiments show that, with only one domainprompt, both methods can effectively reduce word error rates (WER) onout-of-domain TedLium-2 and SPGISpeech datasets. Especially, the deepLLM-fusion has the advantage of better recall of entity and out-of-vocabularywords.

Prompting Large Language Models for Zero-Shot Domain Adaptation in Speech Recognition