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

Recently, self-supervised pretraining has achieved impressive results inend-to-end (E2E) automatic speech recognition (ASR). However, the dominantsequence-to-sequence (S2S) E2E model is still hard to fully utilize theself-supervised pre-training methods because its decoder is conditioned onacoustic representation thus cannot be pretrained separately. In this paper, wepropose a pretrained Transformer (Preformer) S2S ASR architecture based onhybrid CTC/attention E2E models to fully utilize the pretrained acoustic models(AMs) and language models (LMs). In our framework, the encoder is initializedwith a pretrained AM (wav2vec2.0). The Preformer leverages CTC as an auxiliarytask during training and inference. Furthermore, we design a one-cross decoder(OCD), which relaxes the dependence on acoustic representations so that it canbe initialized with pretrained LM (DistilGPT2). Experiments are conducted onthe AISHELL-1 corpus and achieve a $4.6\%$ character error rate (CER) on thetest set. Compared with our vanilla hybrid CTC/attention Transformer baseline,our proposed CTC/attention-based Preformer yields $27\%$ relative CERreduction. To the best of our knowledge, this is the first work to utilize bothpretrained AM and LM in a S2S ASR system.

Improving Hybrid CTC/Attention End-to-end Speech Recognition with Pretrained Acoustic and Language Model