Under review

Can i program my AI avatar to only use my voice?

eyeconsult3 8 years ago updated by Gil Sideman 8 years ago 5

I want to be able to use my voice for my AI avatar so that when it answers, it'll be my voice that users hear, whether I've programmed the answer or not. You got what I'm saying?

Under review

Hi Eyeconsult3 -

The ability you are looking for is referred to as a custom TTS voice. Creating a TTS voice to sound just like a specific person's voice.

There are no good solutions for doing so at a reasonable price - which is why we have not included this feature so far in our product platform.

I'm sorry I could not be more helpful.



The SitePal Team

i also wanted to ask about voice activation so users can verbally submit their questions as opposed to typing. Is there anything like that? Thanks for the prompt reply to my earlier question, by the way.

There are speech recognition solutions out there that can be used in conjunction with SitePal. But this capability is not built-in. Google is one possible provider.

Using the SitePal API function 'sayAIResponse' it is possible to link the output from voice recognition with the input to the SitePal AI bot.

So the bottom line is - Yes, this is possible to do but requires work to integrate a 3rd party solution.

I should perhaps add that there is an emerging standard in HTML5 - which makes speech recognition as easy as pie -

see - http://shapeshed.com/html5-speech-recognition-api/

and - http://stephenwalther.com/archive/2015/01/05/using-html5-speech-recognition-and-text-to-speech

problem is it is not widely supported

see - http://caniuse.com/#feat=speech-recognition

Thanks for the response. I have another question. Is there a way to increase the amount of tts characters that I can use? Right now, I have 900 characters. Can I upgrade to get more characters or something?

900 characters is the maximum technical limit.

If you need longer text to be spoken then you can use our API to play back several strings in sequence (the API automatically queues playback).

Here's a technical example -


Hope this helps.