medkit.audio.transcription

APIs

For accessing these APIs, you may use import like this:

from medkit.audio.transcription import <api_to_import>

Classes:

DocTranscriber(input_label, output_label, ...)	Speech-to-text transcriber generating text documents from audio documents.
TranscribedDocument(text, ...[, anns, ...])	Subclass for TextDocument instances generated by audio transcription.
TranscriberFunction(*args, **kwds)	Protocol for components in charge of the actual speech-to-text transcription to use with DocTranscriber
TranscriberFunctionDescription(name[, config])	Description of a specific instance of a transcriber function (similarly to OperationDescription).

class DocTranscriber(input_label, output_label, transcriber_func, attrs_to_copy=None, uid=None)

Speech-to-text transcriber generating text documents from audio documents.

For each text document, all audio segments with a specific label are converted into text segments and regrouped in a corresponding new text document. The text of each segment is concatenated to form the full raw text of the new document.

Generated text documents are instances of TranscribedDocument (subclass of TextDocument) with additional info such as the identifier of the original audio document and a mapping between audio spans and text spans.

Methods :func: create_text_segment() and :func: augment_full_text_for_next_segment() can be overridden to customize how the text segments are created and how they are concatenated to form the full text.

The actual transcription task is delegated to a TranscriberFunction that must be provided.

Parameters

• input_label (str) – Label of audio segments that should be transcribed.

• output_label (str) – Label of generated text segments.

• transcriber_func (TranscriberFunction) – Transcription component in charge of actually transforming each audio signal into text.

• attrs_to_copy (Optional[List[str]]) – Labels of attributes that should be copied from the original audio segments to the transcribed text segments.

• uid (str) – Identifier of the transcriber.

Methods:

augment_full_text_for_next_segment(...)	Append intermediate joining text to full text before the next segment is concatenated to it.
run(audio_docs)	Return a transcribed text document for each document in audio_docs

run(audio_docs)

Return a transcribed text document for each document in audio_docs

Parameters

audio_docs (List[AudioDocument]) – Audio documents to transcribe

Return type

List[TranscribedDocument]

Returns

List[TranscribedDocument] – Transcribed text documents (once per document in audio_docs)

augment_full_text_for_next_segment(full_text, segment_text, audio_segment)

Append intermediate joining text to full text before the next segment is concatenated to it. Override for custom behavior.

Return type

str

class TranscriberFunction(*args, **kwds)

Protocol for components in charge of the actual speech-to-text transcription to use with DocTranscriber

Methods:

transcribe(audios)

Convert audio buffers into strings by performing speech-to-text.

transcribe(audios)

Convert audio buffers into strings by performing speech-to-text.

Parameters

audios (List[AudioBuffer]) – Audio buffers to converted

Return type

List[str]

Returns

List[str] – Text transcription for each buffer in audios

class TranscriberFunctionDescription(name, config=<factory>)

Description of a specific instance of a transcriber function (similarly to OperationDescription).

Parameters

• name (str) – The name of the transcriber function (typically the class name).

• config (Dict[str, Any]) – The specific configuration of the instance.

class TranscribedDocument(text, text_spans_to_audio_spans, audio_doc_id, anns=None, metadata=None, uid=None)

Subclass for TextDocument instances generated by audio transcription.

Variables

• uid (str) – Document identifier.

• text – The full transcribed text.

• text_spans_to_audio_spans (Dict[medkit.core.text.span.Span, medkit.core.audio.span.Span]) – Mapping between text characters spans in this document and corresponding audio spans in the original audio.

• audio_doc_id (Optional[str]) – Id of the original AudioDocument that was transcribed, if known.

• anns (medkit.core.text.annotation_container.TextAnnotationContainer) – Annotations of the document.

• metadata (Dict[str, Any]) – Document metadata.

• raw_segment (medkit.core.text.annotation.Segment) – Auto-generated segment containing the raw full transcribed text.

Methods:

from_dict(doc_dict)	Create a TranscribedDocument from a dict
get_containing_audio_spans(text_ann_spans)	Return the audio spans used to transcribe the text referenced by a text annotation.

get_containing_audio_spans(text_ann_spans)

Return the audio spans used to transcribe the text referenced by a text annotation.

For instance, if the audio ranging from 1.0 to 20.0 seconds is transcribed to some text ranging from character 10 to 56 in the transcribed document, and then a text annotation is created referencing the span 15 to 25, then the containing audio span will be the one ranging from 1.0 to 20.0 seconds.

Note that some text annotations maybe be contained in more that one audio spans.

Parameters

text_ann_spans (List[AnySpan]) – Text spans of a text annotation referencing some characters in the transcribed document.

Return type

List[Span]

Returns

List[AudioSpan] – Audio spans used to transcribe the text referenced by the spans of text_ann.

classmethod from_dict(doc_dict)

Create a TranscribedDocument from a dict

Parameters

doc_dict (Dict[str, Any]) – A dictionary from a serialized TranscribedDocument as generated by to_dict()

Return type

Self

Subpackages / Submodules

medkit.audio.transcription.doc_transcriber
medkit.audio.transcription.hf_transcriber_function	This module needs extra-dependencies not installed as core dependencies of medkit.
medkit.audio.transcription.sb_transcriber_function	This module needs extra-dependencies not installed as core dependencies of medkit.
medkit.audio.transcription.transcribed_document