SpeechRecognition.md

SpeechRecognition

Namespace: SpawnDev.SpawnJS.JSObjects.Speech
Inheritance: SpawnJSObject -> EventTarget -> SpeechRecognition
MDN Reference: SpeechRecognition - MDN

The SpeechRecognition interface of the Web Speech API is the controller for the recognition service. It handles the SpeechRecognitionEvent sent from the recognition service. Use it to transcribe speech from the user's microphone.

Constructors

Signature	Description
SpeechRecognition()	Creates a new SpeechRecognition instance.
SpeechRecognition(SpawnJSObjectReference _ref)	Deserialization constructor.

Properties

Property	Type	Access	Description
Continuous	bool	get/set	If true, results are returned continuously (not just once).
InterimResults	bool	get/set	If true, interim (non-final) results are returned.
Lang	string	get/set	The language for recognition (BCP 47, e.g., "en-US").
MaxAlternatives	int	get/set	Maximum number of alternative transcriptions. Default: 1.
Grammars	SpeechGrammarList	get/set	The grammars for the recognizer.

Methods

Method	Return Type	Description
Start()	void	Starts the recognition service.
Stop()	void	Stops and returns the final result.
Abort()	void	Stops without returning a result.

Events

Event	Type	Description
OnResult	ActionEvent<SpeechRecognitionEvent>	Fired when a result is available.
OnEnd	ActionEvent<Event>	Fired when the service disconnects.
OnStart	ActionEvent<Event>	Fired when recognition starts.
OnError	ActionEvent<SpeechRecognitionErrorEvent>	Fired on error.
OnNoMatch	ActionEvent<SpeechRecognitionEvent>	Fired when no match is found.
OnAudioStart	ActionEvent<Event>	Fired when audio capture starts.
OnAudioEnd	ActionEvent<Event>	Fired when audio capture ends.
OnSoundStart	ActionEvent<Event>	Fired when sound is detected.
OnSoundEnd	ActionEvent<Event>	Fired when sound stops.
OnSpeechStart	ActionEvent<Event>	Fired when speech is detected.
OnSpeechEnd	ActionEvent<Event>	Fired when speech stops.

Example

using var recognition = new SpeechRecognition();
recognition.Lang = "en-US";
recognition.Continuous = true;
recognition.InterimResults = true;

// Subscribe using named methods (required for proper cleanup)
recognition.OnResult += Recognition_OnResult;
recognition.OnError += Recognition_OnError;

recognition.Start();

// Clean up event handlers before disposal
recognition.OnResult -= Recognition_OnResult;
recognition.OnError -= Recognition_OnError;

void Recognition_OnResult(SpeechRecognitionEvent e)
{
    // Process results
    Console.WriteLine("Speech recognized");
}

void Recognition_OnError(SpeechRecognitionErrorEvent e)
{
    Console.WriteLine($"Error: {e.Error}");
}