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

We present an analysis of semi-supervised acoustic and language modeltraining for English-isiZulu code-switched ASR using soap opera speech.Approximately 11 hours of untranscribed multilingual speech was transcribedautomatically using four bilingual code-switching transcription systemsoperating in English-isiZulu, English-isiXhosa, English-Setswana andEnglish-Sesotho. These transcriptions were incorporated into the acoustic andlanguage model training sets. Results showed that the TDNN-F acoustic modelsbenefit from the additional semi-supervised data and that even betterperformance could be achieved by including additional CNN layers. Using theseCNN-TDNN-F acoustic models, a first iteration of semi-supervised trainingachieved an absolute mixed-language WER reduction of 3.4%, and a further 2.2%after a second iteration. Although the languages in the untranscribed data wereunknown, the best results were obtained when all automatically transcribed datawas used for training and not just the utterances classified asEnglish-isiZulu. Despite reducing perplexity, the semi-supervised languagemodel was not able to improve the ASR performance.

Semi-supervised acoustic and language model training for English-isiZulu code-switched speech recognition