Open yml-bit opened 3 years ago
Do you mean that you want to compute I-vectors without using Kaldi?
Yes, I think so.Kaldi is highly integrated, making it difficult to read and understand the Ivector code.Do you have any suggestions for the independence of ivector?
------------------ 原始邮件 ------------------ 发件人: "Srikanth M @.>; 发送时间: 2021年3月11日(星期四) 下午4:42 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [idiap/kaldi-ivector] extract ivector model from kaldi (#6)
Do you mean that you want to compute I-vectors without using Kaldi?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
I'm not aware of something like that (mostly because I didn't look for it). But this is something that might soon come here.
Ok,thanks a lot.
------------------ 原始邮件 ------------------ 发件人: "Srikanth M @.>; 发送时间: 2021年3月11日(星期四) 下午5:42 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [idiap/kaldi-ivector] extract ivector model from kaldi (#6)
I'm not aware of something like that (mostly because I didn't look for it). But this is something that might soon come here.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
Hello, author.Have you considered separating Ivector from Kaldi?When I was trying to separate the ivector from Kaldi, I found that the calculation of statistics and the extraction of ivector depended on base/util/ GMM/HMM, etc. I got confused during the debugging process.Do you have any good methods or suggestions for this?