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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

With the development of deep learning, automatic speech recognition (ASR) hasmade significant progress. To further enhance the performance, revisingrecognition results is one of the lightweight but efficient manners. Variousmethods can be roughly classified into N-best reranking methods and errorcorrection models. The former aims to select the hypothesis with the lowesterror rate from a set of candidates generated by ASR for a given input speech.The latter focuses on detecting recognition errors in a given hypothesis andcorrecting these errors to obtain an enhanced result. However, we observe thatthese studies are hardly comparable to each other as they are usually evaluatedon different corpora, paired with different ASR models, and even use differentdatasets to train the models. Accordingly, we first concentrate on releasing anASR hypothesis revising (HypR) dataset in this study. HypR contains severalcommonly used corpora (AISHELL-1, TED-LIUM 2, and LibriSpeech) and provides 50recognition hypotheses for each speech utterance. The checkpoint models of theASR are also published. In addition, we implement and compare several classicand representative methods, showing the recent research progress in revisingspeech recognition results. We hope the publicly available HypR dataset canbecome a reference benchmark for subsequent research and promote the school ofresearch to an advanced level.

HypR: A comprehensive study for ASR hypothesis revising with a reference corpus