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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Fundamental modelling differences between hybrid and end-to-end (E2E)automatic speech recognition (ASR) systems create large diversity andcomplementarity among them. This paper investigates multi-pass rescoring andcross adaptation based system combination approaches for hybrid TDNN andConformer E2E ASR systems. In multi-pass rescoring, state-of-the-art hybridLF-MMI trained CNN-TDNN system featuring speed perturbation, SpecAugment andBayesian learning hidden unit contributions (LHUC) speaker adaptation was usedto produce initial N-best outputs before being rescored by the speaker adaptedConformer system using a 2-way cross system score interpolation. In crossadaptation, the hybrid CNN-TDNN system was adapted to the 1-best output of theConformer system or vice versa. Experiments on the 300-hour Switchboard corpussuggest that the combined systems derived using either of the two systemcombination approaches outperformed the individual systems. The best combinedsystem obtained using multi-pass rescoring produced statistically significantword error rate (WER) reductions of 2.5% to 3.9% absolute (22.5% to 28.9%relative) over the stand alone Conformer system on the NIST Hub5'00, Rt03 andRt02 evaluation data.

Two-pass Decoding and Cross-adaptation Based System Combination of End-to-end Conformer and Hybrid TDNN ASR Systems