The importance of phase in speech enhancement

By Kuldip K. Paliwal, Kamil K. Wójcicki, Benjamin J. Shannon. 2011

439 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Computational Mechanics, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Computational Mechanics

In Signal Processing

Citations in the corpus

Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis (2023, background): “Furthermore, the literature does not provide a definitive answer regarding the perceptual importance of phase-related information in speech (Wang & Lim, 1982; Paliwal et al., 2011).”
Time-Frequency Masking in the Complex Domain for Speech Dereverberation and Denoising (2017, positive): “This is important since it means that when it is applied to the reverberant speech, both the magnitude and phase are enhanced, which is crucial for speech quality [31].”
SEGAN: Speech Enhancement Generative Adversarial Network (2017, background): “However, further studies [18] show that signiﬁcant improvements of speech quality are possible, especially when a clean phase spectrum is known.”
Complex Ratio Masking for Monaural Speech Separation (2015, background): “Paliwal et al. combine the noisy magnitude response with the oracle (i.e. clean) phase, non-oracle (i.e. noisy) phase, and enhanced phase where mismatched shorttime Fourier transform (STFT) analysis windows are used to extract the magnitude and phase spectra.”
Deep Learning for Environmentally Robust Speech Recognition (2017, methodology): “However, both the IBM and IRM-based approaches for calculating the target masks simply ignore the distorted phase information, even though it has been shown to be helpful for speech enhancement [93].”
TCNN: Temporal Convolutional Neural Network for Real-time Speech Enhancement in the Time Domain (2019, positive): “Some studies in the past have demonstrated that the phase is necessary for better speech quality, especially in low signal-tonoise ratio (SNR) conditions [8].”
PHASEN: A Phase-and-Harmonics-Aware Speech Enhancement Network (2019, background): “Later, research (Paliwal, W´ojcicki, and Shannon 2011) reveals that phase plays an important role in speech quality and intel-ligibility.”
Raw waveform-based speech enhancement by fully convolutional networks (2017, background): “Several recent studies have revealed the importance of phase when spectrograms are resynthesized back into timedomain waveforms [17, 18].”
Phase Processing for Single-Channel Speech Enhancement: History and recent advances (2015, background): “Therefore, researchers started reinvestigating the role of the STFT phase for speech intelligibility and quality [14], [15].”
Learning Complex Spectral Mapping with Gated Convolutional Recurrent Networks for Monaural Speech Enhancement (2019, background): “[23], however, shows that accurate phase estimation can considerably improve both objective and subjective speech quality, especially when the analysis window for phase spectrum computation is carefully selected.”
Speech dereverberation and denoising using complex ratio masks (2017, background): “, however, shows that the phase response is important for improving the perceptual quality of noisy speech [9].”
End-to-End Waveform Utterance Enhancement for Direct Evaluation Metrics Optimization by Fully Convolutional Neural Networks (2017, background): “However, several recent studies have revealed the importance of phase to speech quality when speech is resynthesized back into time-domain waveforms [26], [46], [47].”
Speech Enhancement of Noisy and Reverberant Speech for Text-to-Speech (2018, background): “Phase distortions caused by additive noise are perceptually relevant to humans [57].”
Speech Enhancement with a Generative Adversarial Network (2019, methodology): “That was questioned by Paliwal et al. (2011), which showed that the quality of the enhanced speech could be further improved by including enhancement of the phase spectrum.”
PhaseNet: Discretized Phase Modeling with Deep Neural Networks for Audio Source Separation (2018, background): “However, recent works have shown that estimating phase also improves the perceptual quality and the separation performance [15–17].”
Sergan: Speech Enhancement Using Relativistic Generative Adversarial Networks with Gradient Penalty (2019, background): “Since speech quality can be signiﬁcantly improved when the clean phase spectrum is known [8], it is worthwhile exploring speech enhancement techniques which preserve phase information.”