A astute colleague of mine, Ulf Björklund has written a nice speech recognition application controlling a few JQueryUI widgets over Websockets, it it based on the websocket server of mine. Ulf idea and demo is quite interesting an brings a lot of nice ideas to my mind. We recorded a little demo that you will find below.
JQueryUI Widgets (WebUI) controlled by speech over WebSockets!
Ulf’s client is based on a Windows application that uses the .NET libraries for speech recognition , those are interpreted and sent to the socketserver and the webpages to control by using speech. In addition to that Ulf added some other nice features. Have a look at the short video demonstration.
As I don’t can keep my hands of did some testing replacing the “Windows client” with Google Chromes build in speech recognition by using
<input type="text" x-webkit-speech />
I made a quick and dirty client that you can try.
1. Start the http://kampanj.xlent.se/sockets/speech/client.htm , just make sure you start Chrome by passing --enable-speech-input switch to Chrome.exe when staring the browser. And you will se a microphone within the “textbox” meaning that you are in input speech mode …
2. As step to you access the WebUI shown in Ulfs video at http://kampanj.xlent.se/sockets/speech/ , make sure you see both windows at the same time (each url started in separate instances of Chrome browsers) .
3. As my demo is very quick I just tested one of the commands Ulf added ; try say “show dialog” .
Magnus Thor
0 kommentarer:
Skicka en kommentar