Abstract: In this work, we propose CleanMel, a single-channel Mel-spectrogram denoising and dereverberation network for improving both speech quality and automatic speech recognition (ASR) performance ...

Abstract: Acoustic features play an important role in improving the quality of the synthesised speech. Currently, the Mel spectrogram is a widely employed acoustic feature in most acoustic models.

A violin plot is an easy to read substitute for a box plot that replaces the box shape with a kernel density estimate of the data, and optionally overlays the data points itself. The original boxplot ...