Speech To Text Android Extension

[Fernando]

Are you using latest android studio JDK?, or OpenJDK?

this error---> java.lang.IllegalArgumentException: Unsupported class file major version 60

show me you are using java from oracle or an outdated java SDK

[VBEinc]

Hi, just looked the code quickly and perform some compilation changes , result it's working for me;
I must change some menu options on C++ part, but first you must test the extension if it's ok i will do the rest to release it;

.mfa example is inside strange thing anyway the scene contain a counter object to receive texte from extension otherwise you got black area on text object;

Be sure have the latest android studio and the sdk ... etc

Regards

Well, it works somewhat well.
I found an issue with your TTS and STT object on the same app.
Basically, when you have both. The only text it wants to retrieve from the STT object is the " Listening " from I assume the STT Object's Default code.
I noticed this on the original STT you released too, when I added the TTS Object to the mix.

Basically, I set it up so that once STT was FINISHED LISTENING the TTS Object would Speak what the STT Object Heard. The only thing that the TTS Object would say is " Listening "

I emphasize the Capitalization as Important as it is Exactly typed out like this. I assume this is in your code somewhere.

[VBEinc]

This is what your Example does with TTS and STT in the same app.
I actually said," HELLO " which it seen and typed out but quickly changed it to "Listening" and Said," Listening" .
I added nothing to say," Listening" anywhere. It just magically pops up. This is why I assume it is in your Code Somewhere.
Screenshot_20221202-035747.jpg

I also noticed when referring anything to to STT Object it shows TTS in the Code.
TTSGetText$( "Android STT" )
Which this shows up when using the STT to Get Text.
I was expecting to see STTGetText#("Android STT") instead.


Last but not Least.
The TTS and STT objects look almost Identical.
Which,( personal preference ) is a pain as I keep clicking the wrong one or selecting the wrong one for a Project.
LOOKTHESAME.jpg

So, if you want to use these. Feel Free! I made some new ones for you.
TTS
TTS.png

STT
STT.png

Thank you, for the Hard work so far. These objects are awesome! Can't wait till they both work perfectly!

[Gigatron]

Don't worry, sooner or later I will finish both expansions so that they work perfectly.

[VBEinc]

Don't worry, sooner or later I will finish both expansions so that they work perfectly.

Well, I look Forward to it when you do. My Wallet is waiting on the Second one and I am still waiting on the first one I bought.
Both are Very Impressive for Fusion to begin with!

Thanks for your Hard work

[VBEinc]

Well, I got it working with a counter.
The whole hidden "Listening" text is now not showing or being spoken.
Nice work bud!
I sent you a message with the working .apk with both the STT and TTS working together.
Screenshot_20221202-225750.jpg

[VBEinc]

Ugh, Never mind. The event " IS STT FINISHED " is back to carrying on again.
Basically the first go works fine but the second try carries the event over from above.
Again the event of " IS STT FINISHED " is acting like an always event instead of a single trigger event like" HAS SAMPLE STOPPED ".

[VBEinc]

Not that it is important right now, but I have some ideas for this object I believe would make more people want it.

1) Give the ability to link the listening time to the user via a counter link or say an on / off button via an Object.
I.E. User clicks a button to start listening and the amount of time the listening happens is determined by a Linked counter, that counts down or up and stops at a specific value or an on / off button created by the user with and object. Object facing left is off / object facing right is on. This would allow Developers to create apps for Transcribing audio. Hello! $$$

2) Sounds or No Sounds
I.E. Listening start sound and ending sounds linked to samples in the app instead of what sounds like the default Google Android sounds, could be changed by user. -or- User chooses graphics to display the two events.

3) STT IS LISTENING
I.E. User could link this to a counter for time.

Just ideas bud. Main goal now though, is getting it running perfect with no counters.

I.E. STT IS FINISHED = Speech to Text has JUST Finished Listening. One Event!

Nevertheless, monumental on your part for what you have so far.
Honestly one of the greatest Objects I have seen for Android yet on Fusion.
The Second is the TTS object with the other issues I mentioned to you months ago.
However, Both are still awesome and I Thank you for putting in the work.
I will be supporting the work when you release the Object as complete like I did with the TTS object ( with issues still ) . lol

Great work so far!

[VBEinc]

Hi,
Late on night i have finished at 85% SST extension, you can speak and the extension write the sentence on
an edit box. This extension will go to clickstore;

Stay tuned

Well, I managed to get it to Kind of work now with a Counter to force individual actions.
LINK to BETA STT to TTS app: https://drive.google.com/file/d/1FTI...usp=share_link

What I have found so far with the working app.
1) If you do not say anything. After 6 seconds. The STT Object forces," Listening " as what was heard.
2) If you say something the STT Object does not understand. Again, it forces the ," Listening " as what was heard.

Also
The Event Function of STT FAILED is almost impossible to get to register the event. I even linked it to the counter as number 5 with a 1 second delay as this 1 second delay is why the STT Object is now working in this app.

Basically, this is how I set it up.
1) When you press the " Start Listening " Button. The counter goes to 1
2) This starts the listening process for the given time of the STT Object ( Which I hope you add the ability to control this time as I counted 6 seconds before the app stops listening ).
3) Once the app has heard what was said. The event is to reset the timer to 0 and set the counter to 2
4) After 1 second has passed, the String is set to what was Heard and the timer is set to 0 again. Last the Counter is set to 3
5) Once the counter reaches 3 The TTS Engine kicks in on a " RUN THIS EVENT ONCE WHEN LOOPS " basis and speaks the string displayed. The Counter is now set to 4

Then the cycle repeats!

Here is the .mfa as well.
LINK: https://drive.google.com/file/d/1T6k...usp=share_link

Currently the Android app version ( LINK ABOVE ) will write out whatever is said as long as you keep talking. This is why Gigatron, I hope you add a function to eliminate the timer or at least link it to a timer in Fusion as this STT Object would be extremely useful for transcribing and other applications without that 6 second no audio timer thing in it. Which by the way, the time is far less if you start talking. So taking a breath will stop the listening function.

Honestly, it would most likely be easier if you just link everything to fusion as you will not have to worry about times and such.
I.E.
User clicks button to listen
User clicks button / a different button to stop listening
User clicks a button to get info heard or could be linked to a timer in fusion

Hope this helps you make this Object amazing. Thanks again!

[Gigatron]

Hi,

Just make a new version of STT; Build with the latest fusion Android runtime, and include .mfa example;
Tested with Android Studio and with installing .apk file on device S21(no sound when clicking Start Listen) S7 edge.

Have Fun (hope it's working for you) ;