我觉得现在象mind一样的人很多
这种软件也就是帮忙采访中记忆的
准确度不是那么严格

引用:

原帖由 aeolus 于 2008-1-29 09:44 AM 发表
it wont help, coz all large vocabulary speech recognition softwares need to be trained to each specific person before they can work to some accuracy
speech recognition has been there since 1960's.

If it is for chinese, the performance of IBM via voice is pretty good since chinese is easier to recognize by voice. I used it to compose my bachelor's thesis 10 years ago.

倒也不一定。我看我同学用的软件，可以很轻易地对付不同的声音，很轻松地记录课堂上的讨论，格式是：
professor: ...
femal: ...
male: ...
male: ...
professor:...

而且准确率极高啊，我还没发现过错误

引用:

原帖由 aeolus 于 2008-1-29 09:44 AM 发表
it wont help, coz all large vocabulary speech recognition softwares need to be trained to each specific person before they can work to some accuracy
speech recognition has been there since 1960's.

引用:

原帖由 thunder 于 2008-1-29 06:03 PM 发表
倒也不一定。我看我同学用的软件，可以很轻易地对付不同的声音，很轻松地记录课堂上的讨论，格式是：
professor: ...
femal: ...
male: ...
male: ...
professor:...

而且准确率极高啊，我还没发现过错误 ...

really? 不是很信耶。他用什么软件啊？
你同学processs过吧？

没任何加工，就是即时记录的，滞后似乎不超过三秒
下次一定问问她软件的名字

引用:

原帖由 chao 于 2008-1-29 05:12 PM 发表

really? 不是很信耶。他用什么软件啊？
你同学processs过吧？