Classes
class	AlignmentConfig
	Configuration for text alignment. Controls edit-distance costs and optional features like repetition detection and noise filtering. More...

class	AlignmentItem
	A single step in the alignment trace, representing one edit operation. More...

class	AlignmentResult
	Result of aligning recognized text against reference text. Contains edit-distance metrics and oral-reading views. More...

class	ASRAudioQualityResult
	Audio quality metrics computed during recognition. Provided as part of the response to help assess recording conditions (noise, clipping, signal level). More...

class	ASRPhone
	A phoneme with timing and pronunciation score. More...

class	ASRResponse
	Wraps a single recognition response, providing access to the recognition result, audio quality metrics, and metadata. Also exposes methods to save response data or queue it for upload to Dashboard. More...

class	ASRResult
	Recognition result containing the recognized text, per-word detail, and confidence. More...

class	ASRWord
	A recognized word with timing information and phoneme-level detail. More...

class	KeenASR
	Main facade for the KeenASR speech recognition plugin. More...

class	TextAligner
	Aligns recognized speech text against a reference text using dynamic programming edit distance. Useful for oral-reading scoring and word-error-rate (WER) computation. More...

class	WordPronunciation
	Specifies an alternative pronunciation for a word in a decoding graph. More...

Enumerations
enum	VadParameter { TimeoutForNoSpeech = 0 , TimeoutEndSilenceForGoodMatch = 1 , TimeoutEndSilenceForAnyMatch = 2 , TimeoutMaxDuration = 3 }
	Voice Activity Detection parameters used to control automatic stop-listening behavior. More...

enum	RecognizerState { NeedsDecodingGraph = 0 , ReadyToListen = 1 , Listening = 2 , FinalProcessing = 3 }
	Recognizer lifecycle states. More...

enum	LogLevel { Debug = 0 , Info = 1 , Warning = 2 }
	Logging verbosity levels for the ASR framework. More...

enum	SpeakingTask { Default = 0 , OralReading = 1 }
	Speaking task type, used when creating decoding graphs to optimize the language model for specific use cases. More...

enum	AlignOp { Match = 0 , Substitution = 1 , Insertion = 2 , Deletion = 3 }
	Alignment operation type for each step in the alignment trace. More...

Enumeration Type Documentation

◆ AlignOp

Alignment operation type for each step in the alignment trace.

Enumerator
Match	Recognized token matches reference token exactly.
Substitution	Recognized token differs from reference token.
Insertion	Extra token in recognized text (not in reference).
Deletion	Missing token from reference (skipped/not spoken).

Logging verbosity levels for the ASR framework.

Enumerator
Debug	Verbose output including internal processing details.
Info	Standard informational messages.
Warning	Only warnings and errors.

Recognizer lifecycle states.

Enumerator
NeedsDecodingGraph	Recognizer is initialized but no decoding graph has been set.
ReadyToListen	A decoding graph is loaded and the recognizer is ready to start listening.
Listening	The recognizer is actively capturing and decoding audio.
FinalProcessing	Audio capture has stopped and the recognizer is computing the final result.

Speaking task type, used when creating decoding graphs to optimize the language model for specific use cases.

Enumerator
Default	General-purpose speech recognition.
OralReading	Optimized for oral reading assessment (read-aloud tasks).

Voice Activity Detection parameters used to control automatic stop-listening behavior.

Enumerator
TimeoutForNoSpeech	Maximum seconds of silence before any speech is detected. If no speech is detected within this interval, listening will stop automatically.
TimeoutEndSilenceForGoodMatch	Seconds of trailing silence required to finalize a good-confidence match.
TimeoutEndSilenceForAnyMatch	Seconds of trailing silence required to finalize any match.
TimeoutMaxDuration	Maximum total duration of a listening session in seconds.