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

In this paper, we present the system submission for the NIST 2018 SpeakerRecognition Evaluation by DKU Speech and Multi-Modal Intelligent InformationProcessing (SMIIP) Lab. We explore various kinds of state-of-the-art front-endextractors as well as back-end modeling for text-independent speakerverifications. Our submitted primary systems employ multiple state-of-the-artfront-end extractors, including the MFCC i-vector, the DNN tandem i-vector, theTDNN x-vector, and the deep ResNet. After speaker embedding is extracted, weexploit several kinds of back-end modeling to perform variability compensationand domain adaptation for mismatch training and testing conditions. The finalsubmitted system on the fixed condition obtains actual detection cost of 0.392and 0.494 on CMN2 and VAST evaluation data respectively. After the officialevaluation, we further extend our experiments by investigating multipleencoding layer designs and loss functions for the deep ResNet system.

The DKU-SMIIP System for NIST 2018 Speaker Recognition Evaluation