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

Most speech enhancement algorithms make use of the short-time Fouriertransform (STFT), which is a simple and flexible time-frequency decompositionthat estimates the short-time spectrum of a signal. However, the duration ofshort STFT frames are inherently limited by the nonstationarity of speechsignals. The main contribution of this paper is a demonstration of speechenhancement and automatic speech recognition in the presence of reverberationand noise by extending the length of analysis windows. We accomplish thisextension by performing enhancement in the short-time fan-chirp transform(STFChT) domain, an overcomplete time-frequency representation that is coherentwith speech signals over longer analysis window durations than the STFT. Thisextended coherence is gained by using a linear model of fundamental frequencyvariation of voiced speech signals. Our approach centers around using asingle-channel minimum mean-square error log-spectral amplitude (MMSE-LSA)estimator proposed by Habets, which scales coefficients in a time-frequencydomain to suppress noise and reverberation. In the case of multiplemicrophones, we preprocess the data with either a minimum variancedistortionless response (MVDR) beamformer, or a delay-and-sum beamformer (DSB).We evaluate our algorithm on both speech enhancement and recognition tasks forthe REVERB challenge dataset. Compared to the same processing done in the STFTdomain, our approach achieves significant improvement in terms of objectiveenhancement metrics (including PESQ---the ITU-T standard measurement for speechquality). In terms of automatic speech recognition (ASR) performance asmeasured by word error rate (WER), our experiments indicate that the STFT witha long window is more effective for ASR.

Enhancement and Recognition of Reverberant and Noisy Speech by Extending Its Coherence