Keywords feature missing

[zkaiser]

[speech-to-text][v2.15.0][Unity 2018.2.0f2] Including keywords to be detected an error occurs : "Failed to enter listening state", and everything is ok if keywords not set

[mediumTaj]

Hi @zkaiser . Can you post sample code you are using? Did you set KeywordsThreshold?

_service.Keywords = new string[] { "IBM", "Watson"};
_service.KeywordsThreshold = 0.5f;

Keywords is working in the ExampleStreaming scene if you add these two lines to the Active setter. If the threshold wasn't set you would see the error:

Error: keywords_threshold must be set if keywords list is not empty.

[zkaiser]

Hi mediumTaj,

I´m using SpeechSandboxStreaming example with this:

_speechToText.RecognizeModel = "es-ES_NarrowbandModel";//para que reconozca en español -- es-ES_NarrowbandModel -- es-ES_BroadbandModel

List keywords = new List();
keywords.Add("ayuda");
keywords.Add("buscandome");
_speechToText.KeywordsThreshold = 0.1f;
_speechToText.StreamMultipart = false;
_speechToText.Keywords = keywords.ToArray();
_speechToText.InactivityTimeout = 600;
_speechToText.DetectSilence = true;
_speechToText.EnableWordConfidence = true;
_speechToText.EnableTimestamps = true;
_speechToText.SilenceThreshold = 0.01f;
_speechToText.MaxAlternatives = 7;//0
_speechToText.WordAlternativesThreshold = 0.2f;
_speechToText.EnableInterimResults = true;
_speechToText.OnError = OnError;
_speechToText.ProfanityFilter = false;
_speechToText.SmartFormatting = true;
_speechToText.SpeakerLabels = false;
_speechToText.WordAlternativesThreshold = null;*/
_speechToText.StartListening(OnRecognize, OnRecognizeSpeaker);

And I´m getting the error I posted before.
I´m getting the same error when trying ExampleStreaming.unity scene adding keywords (with threshold, of course), and the two scenes works ok without keywords....

When running ExampleSpeechToText example I´m getting following error:

https://gateway-lon.watsonplatform.net/speech-to-text/api/v1/recognize?inactivity_timeout=120&keywords=speech&keywords_threshold=0,1&max_alternatives=1&model=en-US_BroadbandModel&profanity_filter=true&smart_formatting=false&speaker_labels=false&timestamps=false&word_confidence=false, ErrorCode: 400, Error: Generic/unknown HTTP error, Response: {
"code_description": "Bad Request",
"code": 400,
"error": "Expected float for parameter "keywords_threshold", value "0,1" cannot be parsed to float."
}

Using following parameters:
List keywords = new List();
keywords.Add("speech");
_service.KeywordsThreshold = 0.1f;
_service.InactivityTimeout = 120;
_service.StreamMultipart = false;
_service.Keywords = keywords.ToArray();
_service.Recognize(HandleOnRecognize, OnFail, _acousticResourceData, _acousticResourceMimeType);

Is there some version of IBM SDK or Unity SDK do you know where this is working??

[mediumTaj]

@zkaiser Can you take a look at this issue #517

It looks like we need to make an update to an external dependency to specify culture in a serialization script.

[mediumTaj]

This should be fixed with #520. Please comment if you have any more issues with this.