Description

The Azure Speech service provides speech to text and text to speech capabilities with a Speech resource. You can transcribe speech to text with high accuracy, produce natural-sounding text to speech voices, translate spoken audio, and use speaker recognition during conversations.

Create custom voices, add specific words to your base vocabulary, or build your own models. Run Speech anywhere, in the cloud or at the edge in containers. It's easy to speech enable your applications, tools, and devices with the Speech CLI, Speech SDK, Speech Studio, or REST APIs.

Use Azure Speech Services API to:

  • Custom Speech: With Custom Speech, you can upload your own data, test and train a custom model, compare accuracy between models, and deploy a model to a custom endpoint. Copy models to other subscriptions if you want colleagues to have access to a model that you built, or if you want to deploy a model to more than one region.
  • Batch transcription: Transcribe audio files as a batch from multiple URLs or an Azure container.
Supported Operations

Speech Services API v3.1

Get Datasets

Gets a list of datasets for the authenticated subscription.

Create Dataset

Uploads and creates a new dataset by getting the data from a specified URL or starts waiting for data blocks to be uploaded.

Get Supported Locales for Datasets

Gets a list of supported locales for datasets.

Upload Dataset From Form

Uploads data and creates a new dataset.

Get Dataset

Gets the dataset identified by the given ID.

Update Dataset

Updates the mutable details of the dataset identified by its ID.

Delete Dataset

Deletes the specified dataset.

Get Uploaded Blocks

Gets the list of uploaded blocks for this dataset.

Upload Data Block

Upload a block of data for the dataset. The maximum size of the block is 8MiB.

Commit Block List

Commit block list to complete the upload of the dataset.

Get Dataset Files

Gets the files of the dataset identified by the given ID.

Get Dataset File

Gets one specific file (identified with fileId) from a dataset (identified with id).

Get Endpoints

Gets the list of endpoints for the authenticated subscription.

Create Endpoint

Creates a new endpoint.

Get Base Model Logs

Gets the list of audio and transcription logs that have been stored when using the default base model of a given language.

Delete All Base Model Endpoint Logs

Deletion process is done asynchronously and can take up to one day depending on the amount of log files.

Get Base Model Endpoint Log

Gets a specific audio or transcription log for the default base model in a given language.

Delete Base Model Endpoint Log

Deletes one audio or transcription log that have been stored when using the default base model of a given language.

Get Supported Locales for Endpoints

Gets a list of supported locales for endpoint creations.

Get Endpoint

Gets the endpoint identified by the given ID.

Update Endpoint

Updates the metadata of the endpoint identified by the given ID.

Delete Endpoint

Deletes the endpoint identified by the given ID.

Get Endpoint Logs

Gets the list of audio and transcription logs that have been stored for a given endpoint.

Delete All Custom Model Endpoint Logs

The deletion process is done asynchronously and can take up to one day depending on the amount of log files.

Get Custom Model Endpoint Log

Gets a specific audio or transcription log for a given endpoint.

Delete Custom Model Endpoint Log

Deletes one audio or transcription log that have been stored for a given endpoint.

Get Evaluations

Gets the list of evaluations for the authenticated subscription.

Create Evaluation

Creates a new evaluation.

Get Supported Locales for Evaluations

Gets a list of supported locales for evaluations.

Get Evaluation

Gets the evaluation identified by the given ID.

Update Evaluation

Updates the mutable details of the evaluation identified by its id.

Delete Evaluation

Deletes the evaluation identified by the given ID.

Get Evaluation Files

Gets the files of the evaluation identified by the given ID.

Get Evaluation File

Gets one specific file (identified with fileId) from an evaluation (identified with id).

Get Health Status

Returns the overall health of the service and optionally of the different subcomponents.

Get Models

Gets the list of custom models for the authenticated subscription.

Create Model

Creates a new model.

Get Base Models

Gets the list of base models for the authenticated subscription.

Get Base Model

Gets the base model identified by the given ID.

Get Base Model Manifest

Returns an manifest for this base model which can be used in an on-premise container.

Get Supported Locales for Models

Gets a list of supported locales for model adaptation.

Get Model

Gets the model identified by the given ID.

Update Model

Updates the metadata of the model identified by the given ID.

Delete Model

Deletes the model identified by the given ID.

Get Model Files

Gets the files of the model identified by the given ID.

Get Model File

Gets one specific file (identified with fileId) from a model (identified with id).

Get Model Manifest

Returns an manifest for this model which can be used in an on-premise container.

Copy Model

This method can be used to copy a model from one location to another. If the target subscription key belongs to a subscription created for another location, the model will be copied to that location. Only adapted models are allowed to copy to another subscription.

Get Projects

Gets the list of projects for the authenticated subscription.

Create Project

Creates a new project.

Get Supported Locales for Projects

Gets the list of supported locales.

Get Project

Gets the project identified by the given ID.

Update Project

Updates the project identified by the given ID.

Delete Project

Deletes the project identified by the given ID.

Get Datasets for Project

Gets the list of datasets for specified project.

Get Endpoints for Project

Gets the list of endpoints for specified project.

Get Evaluations for Project

Gets the list of evaluations for specified project.

Get Models for Project

Gets the list of models for specified project.

Get Transcriptions for Project

Gets the list of transcriptions for specified project.

Get Transcriptions

Gets a list of transcriptions for the authenticated subscription.

Create Transcription

Creates a new transcription.

Get Supported Locales for Transcriptions

Gets a list of supported locales for offline transcriptions.

Get Transcription

Gets the transcription identified by the given ID.

Update Transcription

Updates the mutable details of the transcription identified by its ID.

Delete Transcription

Deletes the specified transcription task.

Get Transcription Files

Gets the files of the transcription identified by the given ID.

Get Transcription File

Gets one specific file (identified with fileId) from a transcription (identified with id).

Get Web Hooks

Gets the list of web hooks for the authenticated subscription.

Create Web Hook

If the property secret in the configuration is present and contains a non-empty string, it will be used to create a SHA256 hash of the payload with the secret as HMAC key. This hash will be set as X-MicrosoftSpeechServices-Signature header when calling back into the registered URL. When calling back into the registered URL, the request will contain a X-MicrosoftSpeechServices-Event header containing one of the registered event types. There will be one request per registered event type. After successfully registering the web hook, it will not be usable until a challenge/response is completed. To do this, a request with the event type challenge will be made with a query parameter called validationToken. Respond to the challenge with a 200 OK containing the value of the validationToken query parameter as the response body. When the challenge/response is successfully completed, the web hook will begin receiving events.

Get Web Hook

Gets the web hook identified by the given ID.

Update Web Hook

If the property secret in the configuration is omitted or contains an empty string, future callbacks won't contain X-MicrosoftSpeechServices-Signature headers. If the property contains a non-empty string, it will be used to create a SHA256 hash of the payload with the secret as HMAC key. This hash will be set as X-MicrosoftSpeechServices-Signature header when calling back into the registered URL. If the URL changes, the web hook will stop receiving events until a challenge/response is completed. To do this, a request with the event type challenge will be made with a query parameter called validationToken. Respond to the challenge with a 200 OK containing the value of the validationToken query parameter as the response body. When the challenge/response is successfully completed, the web hook will begin receiving events.

Delete Web Hook

Deletes the web hook identified by the given ID.

Ping Web Hook

The request body of the POST request sent to the registered web hook URL is of the same shape as in the GET request for a specific hook. The Swagger Schema ID of the model is WebHookV3. The request will contain a X-MicrosoftSpeechServices-Event header with the value ping. If the web hook was registered with a secret it will contain a X-MicrosoftSpeechServices-Signature header with an SHA256 hash of the payload with the secret as HMAC key. The hash is base64 encoded.

Test Web Hook

The payload will be generated from the last entity that would have invoked the web hook. If no entity is present for none of the registered event types, the POST will respond with 204. If a test request can be made, it will respond with 200. The request will contain a X-MicrosoftSpeechServices-Event header with the respective registered event type. If the web hook was registered with a secret it will contain a X-MicrosoftSpeechServices-Signature header with an SHA256 hash of the payload with the secret as HMAC key. The hash is base64 encoded.

Details
Preview

This item is available for early access. It is still in development and may contain experimental features or limitations.

Last Update

4 months ago

Includes
azure-ai-speech-api
Links
Azure AI Speech API Docs
Azure AI Speech Website
Getting started