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/

Zendy Plus

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

Zendy Tools

Zendy Open

This paper describes our submission to Task 1 of the Short-duration SpeakerVerification (SdSV) challenge 2020. Task 1 is a text-dependent speakerverification task, where both the speaker and phrase are required to beverified. The submitted systems were composed of TDNN-based and ResNet-basedfront-end architectures, in which the frame-level features were aggregated withvarious pooling methods (e.g., statistical, self-attentive, ghostVLAD pooling).Although the conventional pooling methods provide embeddings with a sufficientamount of speaker-dependent information, our experiments show that theseembeddings often lack phrase-dependent information. To mitigate this problem,we propose a new pooling and score compensation methods that leverage aCTC-based automatic speech recognition (ASR) model for taking the lexicalcontent into account. Both methods showed improvement over the conventionaltechniques, and the best performance was achieved by fusing all theexperimented systems, which showed 0.0785% MinDCF and 2.23% EER on thechallenge's evaluation subset.

Robust Text-Dependent Speaker Verification via Character-Level Information Preservation for the SdSV Challenge 2020