How to get echo signal based on the far end speech?

microsoft / AEC-Challenge

AEC Challenge

MIT License

378 stars 128 forks source link

How to get echo signal based on the far end speech? #4

Closed YoungJay0612 closed 3 years ago

YoungJay0612 commented 3 years ago

Hello guy, Sorry to bother you. But I have a doubt about the simulation form far end speech to the echo signal. As we know, it is a nonlinear distortion process, so I wonder how to simulate the nonlinear distortion and get the echo signal? Hopelly to hear from you, Thanks so much!

andosa commented 3 years ago

There is no out-of-the box solution for this. You can apply non-linear transformations on the waveform and apply RIR, or learn the NN models of distortion/RIR functions from the data