How to separate the target speech?

[ucasiggcas]

Dear,

Thanks to your great Project,

I have a question now, if I only have the target person's corpus, and how I separate his speech from multi-talker ?

Any suggestion or advice will be okey.

Thx

Lychee

[snsun]

Hi Lychee,

Sorry that I didn't understand "If I only have the target person's courpus?"

Best Sining

Lychee Tsu [email protected] 于2019年4月25日周四 下午5:55写道：

Dear,

Thanks to your great Project,

I have a question now, if I only have the target person's corpus, and how I separate his speech from multi-talker ?

Any suggestion or advice will be okey.

Thx

Lychee

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/snsun/pit-speech-separation/issues/25, or mute the thread https://github.com/notifications/unsubscribe-auth/ADCLQKO4L3XXC6SHR6A5MXTPSF5ZHANCNFSM4HILRPRQ .

[ucasiggcas]

Emmm,

The target's speech file, that is, audio files, others speech can't get in advance,

For example,

I get Xiaoming's speech every day, but one day he talks with other persons, I just want to separate this talks, one is Xiaoming's speech ,another is others SO what can I do to solve the problem ?

Thx

[snsun]

I see. It seems that you are talking speaker dependent separation, right? You only want to extract Xiaoming's voice from the mixed speech?

Lychee Tsu [email protected] 于2019年4月25日周四 下午8:57写道：

Emmm,

The target's speech file, that is, audio files, others speech can't get in advance,

For example,

I get Xiaoming's speech every day, but one day he talks with other persons, I just want to separate this talks, one is Xiaoming's speech ,another is others SO what can I do to solve the problem ?

Thx

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/snsun/pit-speech-separation/issues/25#issuecomment-486661067, or mute the thread https://github.com/notifications/unsubscribe-auth/ADCLQKJQWYIGRBXNJFAK453PSGTB5ANCNFSM4HILRPRQ .

[ucasiggcas]

Yeah,

I just want Xiaoming's voice, so how to deal with it ?

Thanks a lot

[snsun]

Actually，this is also becoming a hot research topic. I haven't worked on this topic, but I can give you two related papers if you are interested in this task. Maybe you can find something you want from these papers and the references. https://ieeexplore.ieee.org/abstract/document/8683243 https://arxiv.org/abs/1904.03792 https://ieeexplore.ieee.org/abstract/document/8682245

Lychee Tsu [email protected] 于2019年4月26日周五 上午10:33写道：

Yeah,

I just want Xiaoming's voice, so how to deal with it ?

Thanks a lot

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/snsun/pit-speech-separation/issues/25#issuecomment-486903951, or mute the thread https://github.com/notifications/unsubscribe-auth/ADCLQKPULB5THAN3JMYMGZLPSJSXXANCNFSM4HILRPRQ .