Sound Demo of ctPuLSE on CHiME-4 Corpus


This webpage provides a sound demo on the test set of the CHiME-4 corpus. For technical details, please refer to the following paper:

Z.-Q. Wang, "ctPuLSE: Close-Talk, and Pseudo-Label Based Far-Field, Speech Enhancement", in submission, 2024.

This demo presents the enhancement results of 10 randomly-selected test signals.
1. 10 randomly-selected test signals in 6-channel input cases (click to jump to the section)
2. 10 randomly-selected test signals in 2-channel input cases (click to jump to the section)
3. 10 randomly-selected test signals in 1-channel input cases (click to jump to the section)
Firefox is recommended.

Six-Channel Input

Utterance ID=F05_447C020T_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_447C020H_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_447C020M_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_446C020T_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_443C0205_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_447C0202_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_447C0205_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_446C020Z_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_446C020J_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_445C020B_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 6ch
SuperM2M 6ch
ctPuLSE 6ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)

Two-Channel Input

Utterance ID=F05_447C020T_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_447C020H_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_447C020M_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_446C020T_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_443C0205_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_447C0202_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_447C0205_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_446C020Z_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_446C020J_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_445C020B_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 2ch
SuperM2M 2ch
ctPuLSE 2ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)

Monaural Input

Utterance ID=F05_447C020T_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_447C020H_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_447C020M_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_446C020T_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F05_443C0205_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_447C0202_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_447C0205_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_446C020Z_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_446C020J_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)
Utterance ID=F06_445C020B_BUS Target Speaker Log-compreesed Power Spectrogram
Mixture
Supervised 1ch
SuperM2M 1ch
ctPuLSE 1ch
Close-talk Mixture
Est. close-talk speech (CTSEnet)