WHAMR!: Noisy and Reverberant Single-Channel Speech Separation

While significant advances have been made with respect to the separation of overlapping speech signals, studies have been largely constrained to mixtures of clean, near anechoic speech, not representative of many real-world scenarios. Although the WHAM! dataset introduced noise to the ubiquitous wsj...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Maciejewski, Matthew, Wichern, Gordon, McQuinn, Emmett, Jonathan Le Roux
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 14.02.2020
Subjects
Datasets
Galling
Machine learning
Performance degradation
Recording
Separation
Speech
Online AccessGet full text

Cover

More Information
Summary:While significant advances have been made with respect to the separation of overlapping speech signals, studies have been largely constrained to mixtures of clean, near anechoic speech, not representative of many real-world scenarios. Although the WHAM! dataset introduced noise to the ubiquitous wsj0-2mix dataset, it did not include reverberation, which is generally present in indoor recordings outside of recording studios. The spectral smearing caused by reverberation can result in significant performance degradation for standard deep learning-based speech separation systems, which rely on spectral structure and the sparsity of speech signals to tease apart sources. To address this, we introduce WHAMR!, an augmented version of WHAM! with synthetic reverberated sources, and provide a thorough baseline analysis of current techniques as well as novel cascaded architectures on the newly introduced conditions.
ISSN:2331-8422