- [Kaldi-developers] Baum-Welch training for HMMs
- [Kaldi-developers] python wrappers for Kaldi
- Kaldi lattices
- Decoding graph construction in Kaldi: A visual walkthrough
-
Normally you shouldn’t have to set it. It helps to encourage the silence
model to eat up more of the data, to avoid it getting modeled
inappropriately by context-dependent phones. It is helpful on some setups,
but not always.
useful kaldi posts
最新推荐文章于 2019-01-08 07:32:10 发布
本文介绍了Kaldi中HMM的Baum-Welch训练方法、Python接口使用、格子(lattice)概念及解码图构建过程等内容。boost-silence参数用于调整静音模型在数据中的占比,避免被上下文相关电话模型不当占用。此参数有时有助于提高某些配置的表现。
1537

被折叠的 条评论
为什么被折叠?



