0

We are working on an asp.net web application that requires some data to be entered by speech.

The user can enter some data using normal user interface however, we want an additional feature where he can enter data by speaking. We can fix the voice commands like to enter "value1" to "data1", user will speak "data1" followed by "value1" (or anything else, that can be fixed later).

I searched over the internet and found that using Microsoft Speech SDK is a solution. We started with some initial implementation and found that it only works with IE and requires a plugin (we were not able to use this plugin in Windows 7, just XP).

Is there any other solution or can SASDK be used for other browsers as well ? Any suggestions would be helpful.

Thanks, Gaurav

4

4 回答 4

1

是的,可以这样做:

  1. 使用 HTML5 的 getUserMedia 捕获音频流
  2. 将其保存到 .WAV,有相应的库
  3. 通过 AJAX 将 .WAV 发送到服务器
  4. 通过 SetInputToWaveFile 方法将 .WAV 输入到 SpeechRecognitionEngine
  5. 获取结果并在 AJAX 调用中返回

一个例子:

http://weblogs.asp.net/ricardoperes/speech-recognition-in-asp-net

于 2014-11-26T15:38:20.280 回答
0

你也许可以使用 Dragon Naturally 来说,他们有一个 SDK 和一个 Internet Explorer 插件。

http://www.nuance.com/naturallyspeaking/pdf/ds_DNS10_SDK_Client.pdf

http://ct.scansoft.com/customerfiles/kbasefiles/3067/wp_DNS_HTML.pdf

我认为它主要是为遍历网页而设计的,但如果有处理自定义语音命令的选项,那么我不明白为什么你不能用命令名称填充文本框。

可能无法提供您需要的东西,但如果您还没有,当然值得一试

于 2011-01-26T14:20:41.310 回答
0

请阅读:语音识别发生了什么?
除非你正在做一个学术项目,否则它现在是不可能的。

于 2011-01-26T13:55:38.173 回答
0

你可以离开这个想法。不可能从任意人那里识别任意值。有一些或多或少成功的项目(如谷歌语音搜索),但它们是专有的、封闭的且不出售。创建这样的系统的成本将估计为数百万美元。

于 2011-01-26T14:11:57.057 回答