azure.cognitiveservices.vision.customvision.training.operations module¶

class azure.cognitiveservices.vision.customvision.training.operations.CustomVisionTrainingClientOperationsMixin[source]¶

Bases: object

create_image_regions(project_id, regions=None, custom_headers=None, raw=False, **operation_config)[source]¶

Create a set of image regions.

This API accepts a batch of image regions, and optionally tags, to update existing images with region information. There is a limit of 64 entries in the batch. If all regions are successful created, 200(OK) status code will be returned. Otherwise, 207 (Multi-Status) status code will be returned and detail status for each region will be listed in the response payload.

Parameters

project_id (str) – The project id.
regions (list[ImageRegionCreateEntry]) –
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageRegionCreateSummary or ClientRawResponse if raw=true

Return type

ImageRegionCreateSummary or ClientRawResponse

Raises

CustomVisionErrorException

create_image_tags(project_id, tags=None, custom_headers=None, raw=False, **operation_config)[source]¶

Associate a set of images with a set of tags.

Parameters

project_id (str) – The project id.
tags (list[ImageTagCreateEntry]) – Image Tag entries to include in this batch.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageTagCreateSummary or ClientRawResponse if raw=true

Return type

ImageTagCreateSummary or ClientRawResponse

Raises

CustomVisionErrorException

create_images_from_data(project_id, image_data, tag_ids=None, custom_headers=None, raw=False, **operation_config)[source]¶

Add the provided images to the set of training images.

This API accepts body content as multipart/form-data and application/octet-stream. When using multipart multiple image files can be sent at once, with a maximum of 64 files. If all images are successful created, 200(OK) status code will be returned. Otherwise, 207 (Multi-Status) status code will be returned and detail status for each image will be listed in the response payload.

Parameters

project_id (str) – The project id.
image_data (Generator) – Binary image data. Supported formats are JPEG, GIF, PNG, and BMP. Supports images up to 6MB.
tag_ids (list[str]) – The tags ids with which to tag each image. Limited to 20.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageCreateSummary or ClientRawResponse if raw=true

Return type

ImageCreateSummary or ClientRawResponse

Raises

CustomVisionErrorException

create_images_from_files(project_id, batch, custom_headers=None, raw=False, **operation_config)[source]¶

Add the provided batch of images to the set of training images.

This API accepts a batch of files, and optionally tags, to create images. There is a limit of 64 images and 20 tags. If all images are successful created, 200(OK) status code will be returned. Otherwise, 207 (Multi-Status) status code will be returned and detail status for each image will be listed in the response payload.

Parameters

project_id (str) – The project id.
batch (ImageFileCreateBatch) – The batch of image files to add. Limited to 64 images and 20 tags per batch.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageCreateSummary or ClientRawResponse if raw=true

Return type

ImageCreateSummary or ClientRawResponse

Raises

CustomVisionErrorException

create_images_from_predictions(project_id, batch, custom_headers=None, raw=False, **operation_config)[source]¶

Add the specified predicted images to the set of training images.

This API creates a batch of images from predicted images specified. There is a limit of 64 images and 20 tags. If all images are successful created, 200(OK) status code will be returned. Otherwise, 207 (Multi-Status) status code will be returned and detail status for each image will be listed in the response payload.

Parameters

project_id (str) – The project id.
batch (ImageIdCreateBatch) – Image, tag ids, and metadata. Limited to 64 images and 20 tags per batch.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageCreateSummary or ClientRawResponse if raw=true

Return type

ImageCreateSummary or ClientRawResponse

Raises

CustomVisionErrorException

create_images_from_urls(project_id, batch, custom_headers=None, raw=False, **operation_config)[source]¶

Add the provided images urls to the set of training images.

This API accepts a batch of urls, and optionally tags, to create images. There is a limit of 64 images and 20 tags. If all images are successful created, 200(OK) status code will be returned. Otherwise, 207 (Multi-Status) status code will be returned and detail status for each image will be listed in the response payload.

Parameters

project_id (str) – The project id.
batch (ImageUrlCreateBatch) – Image urls, tag ids, and metadata. Limited to 64 images and 20 tags per batch.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageCreateSummary or ClientRawResponse if raw=true

Return type

ImageCreateSummary or ClientRawResponse

Raises

CustomVisionErrorException

create_project(name, description=None, domain_id=None, classification_type=None, target_export_platforms=None, custom_headers=None, raw=False, **operation_config)[source]¶

Create a project.

Parameters

name (str) – Name of the project.
description (str) – The description of the project.
domain_id (str) – The id of the domain to use for this project. Defaults to General.
classification_type (str) – The type of classifier to create for this project. Possible values include: ‘Multiclass’, ‘Multilabel’
target_export_platforms (list[str]) – List of platforms the trained model is intending exporting to.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Project or ClientRawResponse if raw=true

Return type

Project or ClientRawResponse

Raises

CustomVisionErrorException

create_tag(project_id, name, description=None, type=None, custom_headers=None, raw=False, **operation_config)[source]¶

Create a tag for the project.

Parameters

project_id (str) – The project id.
name (str) – The tag name.
description (str) – Optional description for the tag.
type (str) – Optional type for the tag. Possible values include: ‘Regular’, ‘Negative’, ‘GeneralProduct’
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Tag or ClientRawResponse if raw=true

Return type

Tag or ClientRawResponse

Raises

CustomVisionErrorException

delete_image_regions(project_id, region_ids, custom_headers=None, raw=False, **operation_config)[source]¶

Delete a set of image regions.

Parameters

project_id (str) – The project id.
region_ids (list[str]) – Regions to delete. Limited to 64.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

delete_image_tags(project_id, image_ids, tag_ids, custom_headers=None, raw=False, **operation_config)[source]¶

Remove a set of tags from a set of images.

Parameters

project_id (str) – The project id.
image_ids (list[str]) – Image ids. Limited to 64 images.
tag_ids (list[str]) – Tags to be deleted from the specified images. Limited to 20 tags.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

delete_images(project_id, image_ids=None, all_images=None, all_iterations=None, custom_headers=None, raw=False, **operation_config)[source]¶

Delete images from the set of training images.

Parameters

project_id (str) – The project id.
image_ids (list[str]) – Ids of the images to be deleted. Limited to 256 images per batch.
all_images (bool) – Flag to specify delete all images, specify this flag or a list of images. Using this flag will return a 202 response to indicate the images are being deleted.
all_iterations (bool) – Removes these images from all iterations, not just the current workspace. Using this flag will return a 202 response to indicate the images are being deleted.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

delete_iteration(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)[source]¶

Delete a specific iteration of a project.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

delete_prediction(project_id, ids, custom_headers=None, raw=False, **operation_config)[source]¶

Delete a set of predicted images and their associated prediction results.

Parameters

project_id (str) – The project id.
ids (list[str]) – The prediction ids. Limited to 64.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

delete_project(project_id, custom_headers=None, raw=False, **operation_config)[source]¶

Delete a specific project.

Parameters

project_id (str) – The project id.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

delete_tag(project_id, tag_id, custom_headers=None, raw=False, **operation_config)[source]¶

Delete a tag from the project.

Parameters

project_id (str) – The project id.
tag_id (str) – Id of the tag to be deleted.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

export_iteration(project_id, iteration_id, platform, flavor=None, custom_headers=None, raw=False, **operation_config)[source]¶

Export a trained iteration.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id.
platform (str) – The target platform. Possible values include: ‘CoreML’, ‘TensorFlow’, ‘DockerFile’, ‘ONNX’, ‘VAIDK’
flavor (str) – The flavor of the target platform. Possible values include: ‘Linux’, ‘Windows’, ‘ONNX10’, ‘ONNX12’, ‘ARM’, ‘TensorFlowNormal’, ‘TensorFlowLite’
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Export or ClientRawResponse if raw=true

Return type

Export or ClientRawResponse

Raises

CustomVisionErrorException

export_project(project_id, custom_headers=None, raw=False, **operation_config)[source]¶

Exports a project.

Parameters

project_id (str) – The project id of the project to export.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ProjectExport or ClientRawResponse if raw=true

Return type

ProjectExport or ClientRawResponse

Raises

CustomVisionErrorException

get_artifact(project_id, path, custom_headers=None, raw=False, callback=None, **operation_config)[source]¶

Get artifact content from blob storage, based on artifact relative path in the blob.

Parameters

project_id (str) – The project id.
path (str) – The relative path for artifact.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
callback (Callable[Bytes, response=None]) – When specified, will be called with each chunk of data that is streamed. The callback should take two arguments, the bytes of the current chunk of data and the response object. If the data is uploading, response will be None.
operation_config – Operation configuration overrides.

Returns

object or ClientRawResponse if raw=true

Return type

Generator or ClientRawResponse

Raises

HttpOperationError

get_domain(domain_id, custom_headers=None, raw=False, **operation_config)[source]¶

Get information about a specific domain.

Parameters

domain_id (str) – The id of the domain to get information about.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Domain or ClientRawResponse if raw=true

Return type

Domain or ClientRawResponse

Raises

CustomVisionErrorException

get_domains(custom_headers=None, raw=False, **operation_config)[source]¶

Get a list of the available domains.

Parameters

custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Domain] or ClientRawResponse

Raises

CustomVisionErrorException

get_exports(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)[source]¶

Get the list of exports for a specific iteration.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Export] or ClientRawResponse

Raises

CustomVisionErrorException

get_image_count(project_id, iteration_id=None, tagging_status=None, filter=None, tag_ids=None, custom_headers=None, raw=False, **operation_config)[source]¶

Get the number of images.

The filtering is on an and/or relationship. For example, if the provided tag ids are for the “Dog” and “Cat” tags, then only images tagged with Dog and/or Cat will be returned.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
tagging_status (str) – The tagging status filter. It can be ‘All’, ‘Tagged’, or ‘Untagged’. Defaults to ‘All’. Possible values include: ‘All’, ‘Tagged’, ‘Untagged’
filter (str) – An expression to filter the images against image metadata. Only images where the expression evaluates to true are included in the response. The expression supports eq (Equal), ne (Not equal), and (Logical and), or (Logical or) operators. Here is an example, metadata=key1 eq ‘value1’ and key2 ne ‘value2’.
tag_ids (list[str]) – A list of tags ids to filter the images to count. Defaults to all tags when null.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

int or ClientRawResponse if raw=true

Return type

int or ClientRawResponse

Raises

CustomVisionErrorException

get_image_performance_count(project_id, iteration_id, tag_ids=None, custom_headers=None, raw=False, **operation_config)[source]¶

Gets the number of images tagged with the provided {tagIds} that have prediction results from training for the provided iteration {iterationId}.

The filtering is on an and/or relationship. For example, if the provided tag ids are for the “Dog” and “Cat” tags, then only images tagged with Dog and/or Cat will be returned.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
tag_ids (list[str]) – A list of tags ids to filter the images to count. Defaults to all tags when null.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

int or ClientRawResponse if raw=true

Return type

int or ClientRawResponse

Raises

CustomVisionErrorException

get_image_performances(project_id, iteration_id, tag_ids=None, order_by=None, take=50, skip=0, custom_headers=None, raw=False, **operation_config)[source]¶

Get image with its prediction for a given project iteration.

This API supports batching and range selection. By default it will only return first 50 images matching images. Use the {take} and {skip} parameters to control how many images to return in a given batch. The filtering is on an and/or relationship. For example, if the provided tag ids are for the “Dog” and “Cat” tags, then only images tagged with Dog and/or Cat will be returned.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
tag_ids (list[str]) – A list of tags ids to filter the images. Defaults to all tagged images when null. Limited to 20.
order_by (str) – The ordering. Defaults to newest. Possible values include: ‘Newest’, ‘Oldest’
take (int) – Maximum number of images to return. Defaults to 50, limited to 256.
skip (int) – Number of images to skip before beginning the image batch. Defaults to 0.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[ImagePerformance] or ClientRawResponse

Raises

CustomVisionErrorException

get_image_region_proposals(project_id, image_id, custom_headers=None, raw=False, **operation_config)[source]¶

Get region proposals for an image. Returns empty array if no proposals are found.

This API will get region proposals for an image along with confidences for the region. It returns an empty array if no proposals are found.

Parameters

project_id (str) – The project id.
image_id (str) – The image id.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageRegionProposal or ClientRawResponse if raw=true

Return type

ImageRegionProposal or ClientRawResponse

Raises

CustomVisionErrorException

get_images(project_id, iteration_id=None, tag_ids=None, tagging_status=None, filter=None, order_by=None, take=50, skip=0, custom_headers=None, raw=False, **operation_config)[source]¶

Get images for a given project iteration or workspace.

This API supports batching and range selection. By default it will only return first 50 images matching images. Use the {take} and {skip} parameters to control how many images to return in a given batch. The filtering is on an and/or relationship. For example, if the provided tag ids are for the “Dog” and “Cat” tags, then only images tagged with Dog and/or Cat will be returned.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
tag_ids (list[str]) – A list of tags ids to filter the images. Defaults to all tagged images when null. Limited to 20.
tagging_status (str) – The tagging status filter. It can be ‘All’, ‘Tagged’, or ‘Untagged’. Defaults to ‘All’. Possible values include: ‘All’, ‘Tagged’, ‘Untagged’
filter (str) – An expression to filter the images against image metadata. Only images where the expression evaluates to true are included in the response. The expression supports eq (Equal), ne (Not equal), and (Logical and), or (Logical or) operators. Here is an example, metadata=key1 eq ‘value1’ and key2 ne ‘value2’.
order_by (str) – The ordering. Defaults to newest. Possible values include: ‘Newest’, ‘Oldest’
take (int) – Maximum number of images to return. Defaults to 50, limited to 256.
skip (int) – Number of images to skip before beginning the image batch. Defaults to 0.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Image] or ClientRawResponse

Raises

CustomVisionErrorException

get_images_by_ids(project_id, image_ids=None, iteration_id=None, custom_headers=None, raw=False, **operation_config)[source]¶

Get images by id for a given project iteration.

This API will return a set of Images for the specified tags and optionally iteration. If no iteration is specified the current workspace is used.

Parameters

project_id (str) – The project id.
image_ids (list[str]) – The list of image ids to retrieve. Limited to 256.
iteration_id (str) – The iteration id. Defaults to workspace.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Image] or ClientRawResponse

Raises

CustomVisionErrorException

get_iteration(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)[source]¶

Get a specific iteration.

Parameters

project_id (str) – The id of the project the iteration belongs to.
iteration_id (str) – The id of the iteration to get.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Iteration or ClientRawResponse if raw=true

Return type

Iteration or ClientRawResponse

Raises

CustomVisionErrorException

get_iteration_performance(project_id, iteration_id, threshold=None, overlap_threshold=None, custom_headers=None, raw=False, **operation_config)[source]¶

Get detailed performance information about an iteration.

Parameters

project_id (str) – The id of the project the iteration belongs to.
iteration_id (str) – The id of the iteration to get.
threshold (float) – The threshold used to determine true predictions.
overlap_threshold (float) – If applicable, the bounding box overlap threshold used to determine true predictions.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

IterationPerformance or ClientRawResponse if raw=true

Return type

IterationPerformance or ClientRawResponse

Raises

CustomVisionErrorException

get_iterations(project_id, custom_headers=None, raw=False, **operation_config)[source]¶

Get iterations for the project.

Parameters

project_id (str) – The project id.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Iteration] or ClientRawResponse

Raises

CustomVisionErrorException

get_project(project_id, custom_headers=None, raw=False, **operation_config)[source]¶

Get a specific project.

Parameters

project_id (str) – The id of the project to get.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Project or ClientRawResponse if raw=true

Return type

Project or ClientRawResponse

Raises

CustomVisionErrorException

get_projects(custom_headers=None, raw=False, **operation_config)[source]¶

Get your projects.

Parameters

custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Project] or ClientRawResponse

Raises

CustomVisionErrorException

get_tag(project_id, tag_id, iteration_id=None, custom_headers=None, raw=False, **operation_config)[source]¶

Get information about a specific tag.

Parameters

project_id (str) – The project this tag belongs to.
tag_id (str) – The tag id.
iteration_id (str) – The iteration to retrieve this tag from. Optional, defaults to current training set.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Tag or ClientRawResponse if raw=true

Return type

Tag or ClientRawResponse

Raises

CustomVisionErrorException

get_tagged_image_count(project_id, iteration_id=None, tag_ids=None, custom_headers=None, raw=False, **operation_config)[source]¶

Gets the number of images tagged with the provided {tagIds}.

The filtering is on an and/or relationship. For example, if the provided tag ids are for the “Dog” and “Cat” tags, then only images tagged with Dog and/or Cat will be returned.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
tag_ids (list[str]) – A list of tags ids to filter the images to count. Defaults to all tags when null.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

int or ClientRawResponse if raw=true

Return type

int or ClientRawResponse

Raises

CustomVisionErrorException

get_tagged_images(project_id, iteration_id=None, tag_ids=None, order_by=None, take=50, skip=0, custom_headers=None, raw=False, **operation_config)[source]¶

Get tagged images for a given project iteration.

This API supports batching and range selection. By default it will only return first 50 images matching images. Use the {take} and {skip} parameters to control how many images to return in a given batch. The filtering is on an and/or relationship. For example, if the provided tag ids are for the “Dog” and “Cat” tags, then only images tagged with Dog and/or Cat will be returned.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
tag_ids (list[str]) – A list of tags ids to filter the images. Defaults to all tagged images when null. Limited to 20.
order_by (str) – The ordering. Defaults to newest. Possible values include: ‘Newest’, ‘Oldest’
take (int) – Maximum number of images to return. Defaults to 50, limited to 256.
skip (int) – Number of images to skip before beginning the image batch. Defaults to 0.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Image] or ClientRawResponse

Raises

CustomVisionErrorException

get_tags(project_id, iteration_id=None, custom_headers=None, raw=False, **operation_config)[source]¶

Get the tags for a given project and iteration.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Tag] or ClientRawResponse

Raises

CustomVisionErrorException

get_untagged_image_count(project_id, iteration_id=None, custom_headers=None, raw=False, **operation_config)[source]¶

Gets the number of untagged images.

This API returns the images which have no tags for a given project and optionally an iteration. If no iteration is specified the current workspace is used.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

int or ClientRawResponse if raw=true

Return type

int or ClientRawResponse

Raises

CustomVisionErrorException

get_untagged_images(project_id, iteration_id=None, order_by=None, take=50, skip=0, custom_headers=None, raw=False, **operation_config)[source]¶

Get untagged images for a given project iteration.

This API supports batching and range selection. By default it will only return first 50 images matching images. Use the {take} and {skip} parameters to control how many images to return in a given batch.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id. Defaults to workspace.
order_by (str) – The ordering. Defaults to newest. Possible values include: ‘Newest’, ‘Oldest’
take (int) – Maximum number of images to return. Defaults to 50, limited to 256.
skip (int) – Number of images to skip before beginning the image batch. Defaults to 0.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[Image] or ClientRawResponse

Raises

CustomVisionErrorException

import_project(token, name=None, custom_headers=None, raw=False, **operation_config)[source]¶

Imports a project.

Parameters

token (str) – Token generated from the export project call.
name (str) – Optional, name of the project to use instead of auto-generated name.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Project or ClientRawResponse if raw=true

Return type

Project or ClientRawResponse

Raises

CustomVisionErrorException

publish_iteration(project_id, iteration_id, publish_name, prediction_id, overwrite=None, custom_headers=None, raw=False, **operation_config)[source]¶

Publish a specific iteration.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id.
publish_name (str) – The name to give the published iteration.
prediction_id (str) – The id of the prediction resource to publish to.
overwrite (bool) – Whether to overwrite the published model with the given name (default: false).
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

bool or ClientRawResponse if raw=true

Return type

bool or ClientRawResponse

Raises

CustomVisionErrorException

query_predictions(project_id, query, custom_headers=None, raw=False, **operation_config)[source]¶

Get images that were sent to your prediction endpoint.

Parameters

project_id (str) – The project id.
query (PredictionQueryToken) – Parameters used to query the predictions. Limited to combining 2 tags.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

PredictionQueryResult or ClientRawResponse if raw=true

Return type

PredictionQueryResult or ClientRawResponse

Raises

CustomVisionErrorException

query_suggested_image_count(project_id, iteration_id, tag_ids=None, threshold=None, custom_headers=None, raw=False, **operation_config)[source]¶

Get count of images whose suggested tags match given tags and their probabilities are greater than or equal to the given threshold. Returns count as 0 if none found.

This API takes in tagIds to get count of untagged images per suggested tags for a given threshold.

Parameters

project_id (str) – The project id.
iteration_id (str) – IterationId to use for the suggested tags and regions.
tag_ids (list[str]) – Existing TagIds in project to get suggested tags count for.
threshold (float) – Confidence threshold to filter suggested tags on.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

dict or ClientRawResponse if raw=true

Return type

dict[str, int] or ClientRawResponse

Raises

CustomVisionErrorException

query_suggested_images(project_id, iteration_id, query, custom_headers=None, raw=False, **operation_config)[source]¶

Get untagged images whose suggested tags match given tags. Returns empty array if no images are found.

This API will fetch untagged images filtered by suggested tags Ids. It returns an empty array if no images are found.

Parameters

project_id (str) – The project id.
iteration_id (str) – IterationId to use for the suggested tags and regions.
query (SuggestedTagAndRegionQueryToken) – Contains properties we need to query suggested images.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

SuggestedTagAndRegionQuery or ClientRawResponse if raw=true

Return type

SuggestedTagAndRegionQuery or ClientRawResponse

Raises

CustomVisionErrorException

quick_test_image(project_id, image_data, iteration_id=None, store=True, custom_headers=None, raw=False, **operation_config)[source]¶

Quick test an image.

Parameters

project_id (str) – The project id.
image_data (Generator) – Binary image data. Supported formats are JPEG, GIF, PNG, and BMP. Supports images up to 6MB.
iteration_id (str) – Optional. Specifies the id of a particular iteration to evaluate against. The default iteration for the project will be used when not specified.
store (bool) – Optional. Specifies whether or not to store the result of this prediction. The default is true, to store.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction or ClientRawResponse

Raises

CustomVisionErrorException

quick_test_image_url(project_id, url, iteration_id=None, store=True, custom_headers=None, raw=False, **operation_config)[source]¶

Quick test an image url.

Parameters

project_id (str) – The project to evaluate against.
url (str) – Url of the image.
iteration_id (str) – Optional. Specifies the id of a particular iteration to evaluate against. The default iteration for the project will be used when not specified.
store (bool) – Optional. Specifies whether or not to store the result of this prediction. The default is true, to store.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction or ClientRawResponse

Raises

CustomVisionErrorException

suggest_tags_and_regions(project_id, iteration_id, image_ids, custom_headers=None, raw=False, **operation_config)[source]¶

Suggest tags and regions for an array/batch of untagged images. Returns empty array if no tags are found.

This API will get suggested tags and regions for an array/batch of untagged images along with confidences for the tags. It returns an empty array if no tags are found. There is a limit of 64 images in the batch.

Parameters

project_id (str) – The project id.
iteration_id (str) – IterationId to use for tag and region suggestion.
image_ids (list[str]) – Array of image ids tag suggestion are needed for. Use GetUntaggedImages API to get imageIds.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

list[SuggestedTagAndRegion] or ClientRawResponse

Raises

CustomVisionErrorException

train_project(project_id, training_type=None, reserved_budget_in_hours=0, force_train=False, notification_email_address=None, selected_tags=None, custom_headers=None, raw=False, **operation_config)[source]¶

Queues project for training.

Parameters

project_id (str) – The project id.
training_type (str) – The type of training to use to train the project (default: Regular). Possible values include: ‘Regular’, ‘Advanced’
reserved_budget_in_hours (int) – The number of hours reserved as budget for training (if applicable).
force_train (bool) – Whether to force train even if dataset and configuration does not change (default: false).
notification_email_address (str) – The email address to send notification to when training finishes (default: null).
selected_tags (list[str]) – List of tags selected for this training session, other tags in the project will be ignored.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Iteration or ClientRawResponse if raw=true

Return type

Iteration or ClientRawResponse

Raises

CustomVisionErrorException

unpublish_iteration(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)[source]¶

Unpublish a specific iteration.

Parameters

project_id (str) – The project id.
iteration_id (str) – The iteration id.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

None or ClientRawResponse if raw=true

Return type

None or ClientRawResponse

Raises

CustomVisionErrorException

update_image_metadata(project_id, image_ids, metadata, custom_headers=None, raw=False, **operation_config)[source]¶

Update metadata of images.

This API accepts a batch of image Ids, and metadata, to update images. There is a limit of 64 images.

Parameters

project_id (str) – The project id.
image_ids (list[str]) – The list of image ids to update. Limited to 64.
metadata (dict[str, str]) – The metadata to be updated to the specified images. Limited to 50 key-value pairs per image. The length of key is limited to 256. The length of value is limited to 512.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

ImageMetadataUpdateSummary or ClientRawResponse if raw=true

Return type

ImageMetadataUpdateSummary or ClientRawResponse

Raises

CustomVisionErrorException

update_iteration(project_id, iteration_id, name, custom_headers=None, raw=False, **operation_config)[source]¶

Update a specific iteration.

Parameters

project_id (str) – Project id.
iteration_id (str) – Iteration id.
name (str) – Gets or sets the name of the iteration.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Iteration or ClientRawResponse if raw=true

Return type

Iteration or ClientRawResponse

Raises

CustomVisionErrorException

update_project(project_id, updated_project, custom_headers=None, raw=False, **operation_config)[source]¶

Update a specific project.

Parameters

project_id (str) – The id of the project to update.
updated_project (Project) – The updated project model.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Project or ClientRawResponse if raw=true

Return type

Project or ClientRawResponse

Raises

CustomVisionErrorException

update_tag(project_id, tag_id, updated_tag, custom_headers=None, raw=False, **operation_config)[source]¶

Update a tag.

Parameters

project_id (str) – The project id.
tag_id (str) – The id of the target tag.
updated_tag (Tag) – The updated tag model.
custom_headers (dict) – headers that will be added to the request
raw (bool) – returns the direct response alongside the deserialized response
operation_config – Operation configuration overrides.

Returns

Tag or ClientRawResponse if raw=true

Return type

Tag or ClientRawResponse

Raises

CustomVisionErrorException