The effectivness of the pretrained model on Far field speech enhancement. #9
Description
Hi,
First of all, I would like to thank you for your amazing work!
Regarding my question, I would like to use your model for a far-field speech enhancement problem where the speaker is speaking far from the microphone. Your pretrained model gives good initial results; however, sometimes when the target speaker is a bit farther from the microphone, I cannot hear them clearly. Additionally, if the speech signal of the far speaker is too reverberated, the generated output using your model does not make the speech content clear, considering that I am working with Arabic speech data ( could be language affect the performance as well..) .
could you please help me by answering my question and, if possible, provide detailed steps on how I could use your pretrained model as a base and fine-tune it for my specific task?
Thank you very much in advance.