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

Speech enhancement model is used to map a noisy speech to a clean speech. Inthe training stage, an objective function is often adopted to optimize themodel parameters. However, in most studies, there is an inconsistency betweenthe model optimization criterion and the evaluation criterion on the enhancedspeech. For example, in measuring speech intelligibility, most of theevaluation metric is based on a short-time objective intelligibility (STOI)measure, while the frame based minimum mean square error (MMSE) betweenestimated and clean speech is widely used in optimizing the model. Due to theinconsistency, there is no guarantee that the trained model can provide optimalperformance in applications. In this study, we propose an end-to-endutterance-based speech enhancement framework using fully convolutional neuralnetworks (FCN) to reduce the gap between the model optimization and evaluationcriterion. Because of the utterance-based optimization, temporal correlationinformation of long speech segments, or even at the entire utterance level, canbe considered when perception-based objective functions are used for the directoptimization. As an example, we implement the proposed FCN enhancementframework to optimize the STOI measure. Experimental results show that the STOIof test speech is better than conventional MMSE-optimized speech due to theconsistency between the training and evaluation target. Moreover, byintegrating the STOI in model optimization, the intelligibility of humansubjects and automatic speech recognition (ASR) system on the enhanced speechis also substantially improved compared to those generated by the MMSEcriterion.

End-to-End Waveform Utterance Enhancement for Direct Evaluation Metrics Optimization by Fully Convolutional Neural Networks