Honored Contributor

Voice SDK (Feedback/Issues)

Forum|Forum|4 years ago
October 27, 2021
100 replies
54547 views

Do you have any feedback and/or issues in regards to the Voice SDK? Use this place to discuss, as we'll have members of the team reviewing this thread!

Read the blog on Presence Platform (of which the Voice SDK is a part of) here: https://developer.oculus.com/blog/introducing-presence-platform-unleashing-mixed-reality-and-natural-interaction-for-oculus-developers/

Show previous replies

L

Loloncio2

New Participant

Hi,

I am making a game for the quest with Unity and for a trial I have a microphone with a button that you have to press and say a specific phrase. For that I am using the Voice SDK but I have a problem, it works in the editor but not once I compile and install the apk, I have the microphone permission so that is not the issue.

My GameObject that acts as a button is this:

Microphone Button

And the Microfono script is like this:

private void OnTriggerEnter(Collider other)
{
    if (other.tag == "dedo" && !triggerActivated)
    {
        triggerActivated = true;
        appVoice.Activate(GetRequestEvents());
     }
}
// Set the events for the Voice
private VoiceServiceRequestEvents GetRequestEvents()
{
    VoiceServiceRequestEvents events = new VoiceServiceRequestEvents();
    events.OnInit.AddListener(OnInit);
    events.OnComplete.AddListener(OnComplete);
    return events;
}
// What happens when the button is pressed
public void OnInit(VoiceServiceRequest request){
    micro.GetComponent<Outline>().enabled = false;
    verde.SetActive(true);
    rojo.SetActive(false);
}
// What happens when the transcription is complete
public void OnComplete(VoiceServiceRequest request){
    triggerActivated = false;
    verde.SetActive(false);
    rojo.SetActive(true);
}

And the GameObject also has a child GameObject with the response script (maybe it could go into the parent) like this:

Response

Any idea why this might work on the editor and not on quest 3? it does not even launch the OnInit event since the first line there should remove an outline that is around the object and it does not work.

J

joostvrbeelding

Up and Coming

I am trying to understand the dictation SDK, but when I build and run the Dictation Sample scene it gives the following build error:

An asset is marked with HideFlags.DontSave but is included in the build:
Asset: 'Packages/com.meta.xr.sdk.voice/Lib/Wit.ai/Resources/witai.png'
Asset name: witai
(You are probably referencing internal Unity data in your build.)
UnityEngine.GUIUtility:ProcessEvent (int,intptr,bool&)

What's going wrong here? Is this a bug in the wit.ai package?

Unity 2022.3.20f1
Build for Meta Quest 2

Thanks!

B

Baltimon

New Participant

Hello everyone,

in my Project the App Voice Experience has no Activation/Deactivation Events Folder. So i cant find the On Start Listening (), Event that is mationt in the Tutorial. is there a new Version of the SDK with new Events ? Maby i instaled a older Version ? I did everything like in the tutorial. What is wrong with my projekt ?

https://developer.oculus.com/documentation/unity/voice-sdk-tutorials-3/

J

joostvrbeelding

Up and Coming

Does anyone have a working version of the TTS Voices sample? I already fixed a couple of bugs, but it still isn't working and there is no documentation on how it should work.

Here's what I did already:
Change the canvas to World Space and move it further away from the camera. Screen space in a VR scene is weird, it means the canvas is stuck to the headset, right in front of our eyes.

Change the CameraRig to OvrCameraRigInteraction, because in the original sample scene there were no pointers to interact with the menu buttons. And changed the ray length to 1000 to make sure it reaches the canvas.

It's really dissapointing that you can't rely on the samples to understand how this complex functionality works.

A

Abhinai13

New Participant

Yup, Trying to get the Voice SDK to work as well. Felt the same pains you mentioned with the samples.

Have you gotten either of the below ones to work?
- Tutorial 1 - Enabling App Voice Experiences with Built-In NLP
- Tutorial - Receive Basic Voice Input through Voice SDK

No luck for me so far. Appreciate any insight on features / tutorials you have gotten to work for sure

J

joostvrbeelding

Up and Coming

Sorry for the late reply, working on several projects in parallel.

I finally got it working, I think it was a combination of the second tutorial you mention and the video from Dilmer Valecillos below, because in my own scene I still had the text 'Listening', which is what is used in the Meta tutorial.

Dilmer's video:
https://www.youtube.com/watch?v=huc6z1RWCrE&ab_channel=DilmerValecillos

I did not use the Event handling, I used Dilmer's Voice intent Controller. And then used the OnFullTranscription listener method, not the events or the voice intent stuff from the understnding viewer.

appVoiceExperience.TranscriptionEvents.OnFullTranscription.AddListener(OnFullTranscription);

    public void OnFullTranscription(string transcriptionText)
    {
        Debug.Log("OnFullTranscription called.");
        //voiceFeedbackText.text = transcriptionText;
        if (transcriptionText.Length > 0)
        {
            calledWord = transcriptionText;
            ... here I put my analysis of what the calledWord was, just string comparisons, ignoring case...
        }
    }

Hope this helps

RiverExplorer

Honored Contributor

Maybe, what is the full path to the sample from the top of the project? What version of the API are you using? Pre-60 is different than the mostly broken 60 ones. There have been several versions. FYI: More than a few of the meta samples don't work. This may be one.

RiverExplorer.US]@gmail.com https://RiverExplorer.games

J

joostvrbeelding

Up and Coming

Great, thanks.

Yeah, I'm using the 64.0.0 version and as you can see from my earlier post this is already the second one that does not work.

Full path: Assets/Samples/Meta - Voice SDK - Immersive Voice Commands/64.0.0/TTS Voices

R

rajeshimissyou2024

New Participant

rajesh

J

joostvrbeelding

Up and Coming

I am running into a strange problem. In my app there is a selection scene and then one of the scenes is the one with voice recognition. When I start the app and from the selection scene go to the voice recognition scene it works fine.

But when i go back to the selection scene and then start the voice recognition scene again the voice is not recognised anymore. And I get this error message:

Anyone know how to fix this?

J

joostvrbeelding

Up and Coming

I created a git issue for this, as I don't see much support here anymore.

https://github.com/wit-ai/wit/issues/2732

S

sanketjadhav4808

New Participant

Hi There, want solution for offline-support in voice sdk as I have implemented VoiceSDK TTS but its working online and in the docs they say by using Preloading TTS Files can use it offline but theres no implementation for Preloading TTS Files on internet , I searched everywhere how to implement the offline Preloading TTS as per docs..

A

ap664

Inspiring

Has anyone encountered "[VSDK ERROR] [TTSWit] Web Stream Error" after calling TTSSpeaker.Speak() for strings of a certain size?

Around ~200 characters or so, it begins giving this error. Shorter strings are working as expected.

Was wondering if this is by design, and/or ways to mitigate this, because it seems like a pretty low cut-off point. Thanks!