Different implementations of "Weighted Prediction Error" for speech dereverberation:
Yoshioka, Takuya, and Tomohiro Nakatani. "Generalization of multi-channel linear prediction methods for blind MIMO impulse response shortening." IEEE Transactions on Audio, Speech, and Language Processing 20.10 (2012): 2707-2720.
This code has been tested with Python 3.6.
Clone the repository. Then install it as follows:
cd nara_wpe
pip install --user -e .
Check the example notebook for further details. If you download the example notebook, you can listen to the audio examples of input and dereverberated output, too.
Since 2017-09-05 a TensorFlow implementation is added to nara_wpe. It is tested with a few test cases against the Numpy implementation.