TASLP2019

The code for multi-channel speech enhancement which will be published in TASLP2019.
The latest version is available in https://github.com/sekiguchi92/SpeechEnhancement.

• FCA is a method for general source separation. In fact, it can be available only for speech enhancement because of the strong initial value dependency.
• MNMF is a general source separation method which integrate NMF-based source model into FCA.
• MNMF-DP is a method which integrates deep speech prior into MNMF, and is for speech enhancement.
• ILRMA is a general source separation method which integrate NMF-based source model into rank-1 spatial model.
• ILRMA-DP is a method which integrates deep speech prior into ILRMA, and is for speech enhancement.

Requirement

• Tested on Python3.6
• numpy
• pickle
• librosa
• soundfile
• progressbar2
• chainer (6.1.0 was tested) (for MNMF-DP, ILRMA-DP)
• cupy (6.1.0 was tested) (for GPU accelaration)

Usage

python3 MNMF_DP.py [input_filename] --gpu [gpu_id]

Input is the multichannel observed signals.
If gpu_id < 0, CPU is used, and cupy is not necessary.