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

Conformer models have achieved state-of-the-art(SOTA) results in end-to-endspeech recognition. However Conformer mainly focuses on temporal modeling whilepays less attention on time-frequency property of speech feature. In this paperwe augment Conformer with ConvNeXt and propose Nextformer structure. We usestacks of ConvNeXt block to replace the commonly used subsampling module inConformer for utilizing the information contained in time-frequency speechfeature. Besides, we insert an additional downsampling module in middle ofConformer layers to make our model more efficient and accurate. We conductexperiments on two opening datasets, AISHELL-1 and WenetSpeech. On AISHELL-1,compared to Conformer baselines, Nextformer obtains 7.3% and 6.3% relative CERreduction in non-streaming and streaming mode respectively, and on a muchlarger WenetSpeech dataset, Nextformer gives 5.0%~6.5% and 7.5%~14.6% relativeCER reduction in non-streaming and streaming mode, while keep the computationalcost FLOPs comparable to Conformer. To the best of our knowledge, the proposedNextformer model achieves SOTA results on AISHELL-1(CER 4.06%) andWenetSpeech(CER 7.56%/11.29%).

Nextformer: A ConvNeXt Augmented Conformer For End-To-End Speech Recognition