search

zhr1201 / Multi-channel-speech-extraction-using-DNN

A tensorflow implementation of my paper Combining beamforming and deep neural networks for multi-channel speech extraction
MIT License
61 stars 40 forks source link

forward beamformer #1

Closed sunjunlishi closed 7 years ago

sunjunlishi commented 7 years ago

do you supply code?

zhr1201 commented 7 years ago

Sorry but I can't upload the original beamforming code because that is part of our project with a company.

However, this part is quite simple and a reference can be found at http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=1164847 with less than 20 lines of code and a little effort tuning the reg factor, sure you will get that to work.

Other beamforming methods like MVDR, even blind source separation, can also serve as the first stage enhancement (but not delay and sum because its directivity is poor) and some of the code can be found at GitHub.

sunjunlishi commented 7 years ago

ok thank you

sunjunlishi commented 6 years ago

i am poor on it,could you supply the code,please.