Re: Speakable Item and SpeechRecognitionServer
Re: Speakable Item and SpeechRecognitionServer
- Subject: Re: Speakable Item and SpeechRecognitionServer
- From: Deivy Petrescu <email@hidden>
- Date: Thu, 10 Dec 2009 14:21:14 -0500
On Dec 10, 2009, at 2:06 PM, Bruce Robertson wrote:
> I got good recognition after finally calibrating; but there is definitely something weird going on with first run/second run when calling SpeechRecognitionServer.
>
> On Dec 9, 2009, at 11:12 PM, Thomas Fischer wrote:
>
>> Bruce,
>>
>> I was trying some scripts along those lines, and found similar ones using Google.
>> I gave up trying because I never got the voice recognition system to work properly, i.e. to recognise one word (any word) and then get the next.
>> The SpeechRecognitionServer seemed to run into some sort of timeout and at most one digit was recognised (but rarely).
>> Calibration of the system didn't improve the results significantly.
>> My impression is that in principle this should work, but my speech recognition system (using the built-in microphone) isn't working properly, I don't know why the sentences for calibration were recognised but the list of words in AppleScript wasn't.
>>
>> Thomas
>>
>> Am 09.12.2009 um 21:48 schrieb Bruce Robertson:
>>
>>>
>>> On Dec 8, 2009, at 6:35 PM, Deivy Petrescu wrote:
>>>
>>>>
>>>> On Dec 8, 2009, at 6:50 PM, Bruce Robertson wrote:
>>>>
>>>>> Just thought I'd try this again in case anybody came up with ideas.
>>>>>
>>>>> I have still not found a way to speak an arbitrary number sequence and have
>>>>> it recognized; or a mixed sequence of fixed and arbitrary data such as:
>>>>>
>>>>> "Find phone 1234"
>>>>>
>>>>> Or:
>>>>>
>>>>> "Find phone 5513"
>>>>>
>>>>>
>>>>
>>>>
>>>> Bruce:
>>>>
>>>> <script>
>>>> set j to "1 2 3 4 5"
>>>> tell application "SpeechRecognitionServer" to set l to listen for j with prompt "Say " & j & " now" giving up after 15
>>>> if l = j then say "it worked"
>>>>
>>>> </script>
>>>>
>>>> I thought I sent it to you a while ago!
>>>>
>>>> Deivy Petrescu
>>>> email@hidden
>>>
>>> How about something like this?
>>>
>>> Can't quite make it work and I don't want to give the full prompt every time, and it seems to short circuit after the first success, but can somebody help expand on this?
>>>
>>> set j to {0,1,2,3,4,5,6,7,8,9}
>>> set x to {}
>>> repeat 4
>>> tell application "SpeechRecognitionServer" to set L to listen for j with prompt "Say a number" giving up after 10
>>> copy L to end of x
>>> end repeat
>>>
>>> _______________________________________________
That is correct.
I can corroborate that all scripts run by SpeechRecognitionServer have a flicker to them.... they work, they don't work!
This is not so in Leopard.
However, one thing I've noticed in SL, displaying commands is the default in SL and it does not work in Leopard.
And, if Leopard is an example, then SpeechRecognitionServer fails after a certain amount of time.
I'd write to Apple to improve this capability in its OS.
I believe that speech is, strangely, better in Snow Leopard.... Go figure!
Deivy Petrescu
email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
AppleScript-Users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
Archives: http://lists.apple.com/archives/applescript-users
This email sent to email@hidden