use-livecode Digest, Vol 171, Issue 44

[Richmond Mathewson]

If we start to unpack your post we can come to a slightly simpler 
conclusion than what you set out as your goal initially.

1. Is your main idea to have a program where the device can "do the 
talking" for the patient?

2. As the "owner" of a 94 year old Mother-in-law who suffers from some 
sort of selective dementia I realise that
pointing out to her that her hallucinations are hallucinations (rather 
than reality) is both an uphill struggle
and causes her a lot of distress: oddly enough she's far happier inside 
the comfort zone of her hallucinations.

If "Granny" suffers from Aphasia she has permanent damage in the Broca's 
Area of her brain, normally due to
a stroke: this cannot be sorted out, unfortunately.

She could, also, suffer from a type of dementia that has similar 
symptoms to aphasia.

Doesn't really matter which; the end result is just as bad and bl**dy 
awful for all concerned.

If you are going for #1 then there is absolutely no need for any 
speech-recognition stuff or comparison
between recorded sounds and those on the device.

Richmond.

On 28/12/17 2:26 am, Peter Reid via use-livecode wrote:
> Hi Marc, Paul, Phil, Rick and Richmond
>
> Thanks for your various thoughts.
>
> To put a bit more flesh on this, here's what I'm developing and why...
>
> In the first instance I'm doing this development for a friend who's grandmother suffers from aphasia (saying completely the wrong word).  The idea is that her family can put together sets of words where each word is spoken by the app whilst displaying a relevant picture and optionally a short video clip illustrating the correct mouth shape when saying the word.  The app displays the word in a very large font with a picture, then the app says the word (with the option to see a mouth shape video clip).  The user responds by trying to say the same word and (the thing I can't do yet!) the app gives the user a percentage score that represents how closely the user matched the sample word. If the user's score is above a variable threshold, the user sees a smiley face, otherwise a sad face!
>
> I tried contacting Mark Smith as suggested by Paul but his email address mark at maseurope.net no longer works.
>
> So far I have everything working apart from the comparison of 2 WAV files, in particular the following is working:
>
> - the app checks for the presence of a micro SD card as the source of a collection of word packs
> - a word pack consists of a collection of words in sound (WAV) and picture (JPG) form, optionally with supporting video clips (MP4)
> - the app lets the user select a word pack at start up and loads the pack into RW file space
> - the user taps a large arrow icon to go forwards/backwards through the chosen list of words
> - for each word, the word is displayed in a large font at the top of the landscape screen, with its associated picture occupying the lower half of the screen
> - as the word is displayed, it is spoken
> - the user can tap a microphone icon and record their attempt at the word
> - currently I simulate the comparison of the 2 WAV files
> - the app gives a percentage score and displays a smiley or sad face accordingly
>
> To be honest, my main aim is to help a friend.  At this stage I'm not looking beyond this, so licensing, etc. is not on the horizon at the moment.
>
> Thanks again
>
> Peter
> --
> Peter Reid
> Loughborough, UK
>
>
> _______________________________________________
> use-livecode mailing list
> use-livecode at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-livecode