Snowdar / asv-subtools

An Open Source Tools for Speaker Recognition
Apache License 2.0
590 stars 135 forks source link

The way to directly get the recognition result. #10

Open SunxrCarryQAQ opened 4 years ago

SunxrCarryQAQ commented 4 years ago

您好,请问一下您的baseline里面包含可以直接输入音频单个音频直接可以解码出识别结果的脚本吗?如果有的话可以告知一下是哪个嘛?如果没有的话希望您能提供一下大致的流程和思路,非常感谢!

Xuran

Snowdar commented 4 years ago

Hi,这种用于应用的程序流是没有现成的。如果你需要,可以将前处理,模型提取以及后端打分等各个处理模块抽取出来合在一起写成一个流程。该流程可直接用kaldi的源命令而不用局限实验时用的批处理脚本。

Sent from my iPhone

On Sep 9, 2020, at 9:21 AM, SunxrCarryQAQ notifications@github.com wrote:

 您好,请问一下您的baseline里面包含可以直接输入音频单个音频输出识别结果的脚本吗?如果有的话可以告知一下是哪个嘛?如果没有的话希望您能提供一下大致的流程和思路,非常感谢!

Xuran

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.