dual-path-RNNs-DPRNNs-based-speech-separation icon indicating copy to clipboard operation
dual-path-RNNs-DPRNNs-based-speech-separation copied to clipboard

multichannel

Open shanhaidexiamo opened this issue 4 years ago • 3 comments

It seems that you add a beamforming class in this model ,does this mean we can use multi channel datasets?

shanhaidexiamo avatar Sep 21 '20 06:09 shanhaidexiamo

Hi, e... i think the method can be used in multi-channel condition, but we haven't tried. Since Yi Luo has some work to apply DPRNN on multi-channel data. You can refer it. Or if you have any suggestions, can also tell us. Thank you so much.

ShiZiqiang avatar Sep 22 '20 03:09 ShiZiqiang

Hi, e... i think the method can be used in multi-channel condition, but we haven't tried. Since Yi Luo has some work to apply DPRNN on multi-channel data. You can refer it. Or if you have any suggestions, can also tell us. Thank you so much.

Hi,thank you for your reply. I will try to implement the multichannel. by the way ,I find that the batch size is always set to 1, does that mean that if I set it largerer than 1, there will be a cuda out of memory runtime error.

shanhaidexiamo avatar Sep 22 '20 04:09 shanhaidexiamo

Hi, e... i think the method can be used in multi-channel condition, but we haven't tried. Since Yi Luo has some work to apply DPRNN on multi-channel data. You can refer it. Or if you have any suggestions, can also tell us. Thank you so much.

Hi,thank you for your reply. I will try to implement the multichannel. by the way ,I find that the batch size is always set to 1, does that mean that if I set it largerer than 1, there will be a cuda out of memory runtime error.

Yes. It depend on the size of your GPU memory.

ShiZiqiang avatar Sep 23 '20 01:09 ShiZiqiang