Software
Complex-valued Neural Beamforming: [code] to reproduce the results of our paper:
L. Pfeifenberger, M. Zöhrer, F. Pernkopf, Deep Complex-valued Neural Beamformers
If you use this implementation, please reference to the paper.
Blind Source Separation and Dereverberation: [code] to reproduce the results of our paper:
L. Pfeifenberger, F. Pernkopf, Blind Speech Separation and Dereverberation using Neural Beamforming
If you use this implementation, please reference to the paper.
Cross-Domain Echo Controller: [code] to reproduce the results of our paper:
L. Pfeifenberger, M. Zöhrer, F. Pernkopf, “Acoustic Echo Cancellation with Cross-Domain Learning“
If you use this implementation, please reference to the paper.
Nonlinear Residual Echo Suppressor: [code] to reproduce the results of our paper:
L. Pfeifenberger and F. Pernkopf, “Nonlinear Residual Echo Suppression Using a Recurrent Neural Network”
If you use this implementation, please reference to the paper.
Prototype of a 6-channel microphone array ReSpeaker Core V2, using the BSSD inference code written in Tensorflow, and optimized with C++ and Eigen: