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

To realize robust end-to-end Automatic Speech Recognition(E2E ASR) underradio communication condition, we propose a multitask-based method to jointtrain a Speech Enhancement (SE) module as the front-end and an E2E ASR model asthe back-end in this paper. One of the advantage of the proposed method is thatthe entire system can be trained from scratch. Different from prior works,either component here doesn't need to perform pre-training and fine-tuningprocesses separately. Through analysis, we found that the success of theproposed method lies in the following aspects. Firstly, multitask learning isessential, that is the SE network is not only learning to produce moreIntelligent speech, it is also aimed to generate speech that is beneficial torecognition. Secondly, we also found speech phase preserved from noisy speechis critical for improving ASR performance. Thirdly, we propose a dual channeldata augmentation training method to obtain further improvement.Specifically,we combine the clean and enhanced speech to train the whole system. We evaluatethe proposed method on the RATS English data set, achieving a relative WERreduction of 4.6% with the joint training method, and up to a relative WERreduction of 11.2% with the proposed data augmentation method.

Multitask-Based Joint Learning Approach To Robust ASR For Radio Communication Speech