CustomVisionTrainingClient Class

Reference

Inheritance: azure.cognitiveservices.vision.customvision.training.operations._custom_vision_training_client_operations.CustomVisionTrainingClientOperationsMixin

CustomVisionTrainingClient

msrest.service_client.SDKClient

CustomVisionTrainingClient

Constructor

CustomVisionTrainingClient(endpoint, credentials)

Parameters

Name	Description
endpoint Required	str Supported Cognitive Services endpoints.
credentials Required	None Subscription credentials which uniquely identify client subscription.

Variables

Name	Description
config	CustomVisionTrainingClientConfiguration Configuration for client.

Methods

close	Close the client if keep_alive is True.
create_image_regions	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.
create_image_tags	Associate a set of images with a set of tags.
create_images_from_data	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.
create_images_from_files	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.
create_images_from_predictions	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.
create_images_from_urls	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.
create_project	Create a project.
create_tag	Create a tag for the project.
delete_image_regions	Delete a set of image regions.
delete_image_tags	Remove a set of tags from a set of images.
delete_images	Delete images from the set of training images.
delete_iteration	Delete a specific iteration of a project.
delete_prediction	Delete a set of predicted images and their associated prediction results.
delete_project	Delete a specific project.
delete_tag	Delete a tag from the project.
export_iteration	Export a trained iteration.
export_project	Exports a project.
get_artifact	Get artifact content from blob storage, based on artifact relative path in the blob.
get_domain	Get information about a specific domain.
get_domains	Get a list of the available domains.
get_exports	Get the list of exports for a specific iteration.
get_image_count	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.
get_image_performance_count	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.
get_image_performances	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.
get_image_region_proposals	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.
get_images	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.
get_images_by_ids	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.
get_iteration	Get a specific iteration.
get_iteration_performance	Get detailed performance information about an iteration.
get_iterations	Get iterations for the project.
get_project	Get a specific project.
get_projects	Get your projects.
get_tag	Get information about a specific tag.
get_tagged_image_count	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.
get_tagged_images	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.
get_tags	Get the tags for a given project and iteration.
get_untagged_image_count	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.
get_untagged_images	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.
import_project	Imports a project.
publish_iteration	Publish a specific iteration.
query_predictions	Get images that were sent to your prediction endpoint.
query_suggested_image_count	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.
query_suggested_images	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.
quick_test_image	Quick test an image.
quick_test_image_url	Quick test an image url.
suggest_tags_and_regions	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.
train_project	Queues project for training.
unpublish_iteration	Unpublish a specific iteration.
update_image_metadata	Update metadata of images. This API accepts a batch of image Ids, and metadata, to update images. There is a limit of 64 images.
update_iteration	Update a specific iteration.
update_project	Update a specific project.
update_tag	Update a tag.

close

Close the client if keep_alive is True.

close() -> None

create_image_regions

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.

create_image_regions(project_id, regions=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
regions	list[ImageRegionCreateEntry] Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageRegionCreateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageRegionCreateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_image_tags

Associate a set of images with a set of tags.

create_image_tags(project_id, tags=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
tags	list[ImageTagCreateEntry] Image Tag entries to include in this batch. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageTagCreateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageTagCreateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_images_from_data

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.

create_images_from_data(project_id, image_data, tag_ids=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
image_data Required	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. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageCreateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageCreateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_images_from_files

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.

create_images_from_files(project_id, batch, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
batch Required	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 Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageCreateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageCreateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_images_from_predictions

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.

create_images_from_predictions(project_id, batch, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
batch Required	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 Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageCreateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageCreateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_images_from_urls

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.

create_images_from_urls(project_id, batch, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
batch Required	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 Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageCreateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageCreateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_project

Create a project.

create_project(name, description=None, domain_id=None, classification_type=None, target_export_platforms=None, export_model_container_uri=None, notification_queue_uri=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
name Required	str Name of the project.
description	str The description of the project. Default value: None
domain_id	str The id of the domain to use for this project. Defaults to General. Default value: None
classification_type	str The type of classifier to create for this project. Possible values include: 'Multiclass', 'Multilabel' Default value: None
target_export_platforms	list[str] List of platforms the trained model is intending exporting to. Default value: None
export_model_container_uri	str The uri to the Azure Storage container that will be used to store exported models. Default value: None
notification_queue_uri	str The uri to the Azure Storage queue that will be used to send project-related notifications. See Storage notifications documentation for setup and message format. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Project, <xref:msrest.pipeline.ClientRawResponse>	Project or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

create_tag

Create a tag for the project.

create_tag(project_id, name, description=None, type=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
name Required	str The tag name.
description	str Optional description for the tag. Default value: None
type	str Optional type for the tag. Possible values include: 'Regular', 'Negative', 'GeneralProduct' Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Tag, <xref:msrest.pipeline.ClientRawResponse>	Tag or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_image_regions

Delete a set of image regions.

delete_image_regions(project_id, region_ids, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
region_ids Required	list[str] Regions to delete. Limited to 64.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_image_tags

Remove a set of tags from a set of images.

delete_image_tags(project_id, image_ids, tag_ids, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
image_ids Required	list[str] Image ids. Limited to 64 images.
tag_ids Required	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 Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_images

Delete images from the set of training images.

delete_images(project_id, image_ids=None, all_images=None, all_iterations=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
image_ids	list[str] Ids of the images to be deleted. Limited to 256 images per batch. Default value: None
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. Default value: None
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. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_iteration

Delete a specific iteration of a project.

delete_iteration(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str The iteration id.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_prediction

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

delete_prediction(project_id, ids, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
ids Required	list[str] The prediction ids. Limited to 64.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_project

Delete a specific project.

delete_project(project_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

delete_tag

Delete a tag from the project.

delete_tag(project_id, tag_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
tag_id Required	str Id of the tag to be deleted.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

export_iteration

Export a trained iteration.

export_iteration(project_id, iteration_id, platform, flavor=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str The iteration id.
platform Required	str The target platform. Possible values include: 'CoreML', 'TensorFlow', 'DockerFile', 'ONNX', 'VAIDK', 'OpenVino'
flavor	str The flavor of the target platform. Possible values include: 'Linux', 'Windows', 'ONNX10', 'ONNX12', 'ARM', 'TensorFlowNormal', 'TensorFlowLite' Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Export, <xref:msrest.pipeline.ClientRawResponse>	Export or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

export_project

Exports a project.

export_project(project_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id of the project to export.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ProjectExport, <xref:msrest.pipeline.ClientRawResponse>	ProjectExport or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_artifact

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

get_artifact(project_id, path, custom_headers=None, raw=False, callback=None, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
path Required	str The relative path for artifact.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
callback	Callable[<xref:Bytes>, <xref: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. Default value: None
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Generator, <xref:msrest.pipeline.ClientRawResponse>	object or ClientRawResponse if raw=true

Exceptions

Type	Description
msrest.exceptions.HttpOperationError

get_domain

Get information about a specific domain.

get_domain(domain_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
domain_id Required	str The id of the domain to get information about.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Domain, <xref:msrest.pipeline.ClientRawResponse>	Domain or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_domains

Get a list of the available domains.

get_domains(custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Domain], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_exports

Get the list of exports for a specific iteration.

get_exports(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str The iteration id.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Export], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_image_count

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.

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

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
tagging_status	str The tagging status filter. It can be 'All', 'Tagged', or 'Untagged'. Defaults to 'All'. Possible values include: 'All', 'Tagged', 'Untagged' Default value: None
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'. Default value: None
tag_ids	list[str] A list of tags ids to filter the images to count. Defaults to all tags when null. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
int, <xref:msrest.pipeline.ClientRawResponse>	int or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_image_performance_count

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.

get_image_performance_count(project_id, iteration_id, tag_ids=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	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. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
int, <xref:msrest.pipeline.ClientRawResponse>	int or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_image_performances

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.

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

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	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. Default value: None
order_by	str The ordering. Defaults to newest. Possible values include: 'Newest', 'Oldest' Default value: None
take	int Maximum number of images to return. Defaults to 50, limited to 256. Default value: 50
skip	int Number of images to skip before beginning the image batch. Defaults to 0. Default value: 0
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[ImagePerformance], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_image_region_proposals

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.

get_image_region_proposals(project_id, image_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
image_id Required	str The image id.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageRegionProposal, <xref:msrest.pipeline.ClientRawResponse>	ImageRegionProposal or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_images

Get images for a given project iteration or workspace.

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)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
tag_ids	list[str] A list of tags ids to filter the images. Defaults to all tagged images when null. Limited to 20. Default value: None
tagging_status	str The tagging status filter. It can be 'All', 'Tagged', or 'Untagged'. Defaults to 'All'. Possible values include: 'All', 'Tagged', 'Untagged' Default value: None
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'. Default value: None
order_by	str The ordering. Defaults to newest. Possible values include: 'Newest', 'Oldest' Default value: None
take	int Maximum number of images to return. Defaults to 50, limited to 256. Default value: 50
skip	int Number of images to skip before beginning the image batch. Defaults to 0. Default value: 0
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Image], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_images_by_ids

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.

get_images_by_ids(project_id, image_ids=None, iteration_id=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
image_ids	list[str] The list of image ids to retrieve. Limited to 256. Default value: None
iteration_id	str The iteration id. Defaults to workspace. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Image], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_iteration

Get a specific iteration.

get_iteration(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The id of the project the iteration belongs to.
iteration_id Required	str The id of the iteration to get.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Iteration, <xref:msrest.pipeline.ClientRawResponse>	Iteration or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_iteration_performance

Get detailed performance information about an iteration.

get_iteration_performance(project_id, iteration_id, threshold=None, overlap_threshold=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The id of the project the iteration belongs to.
iteration_id Required	str The id of the iteration to get.
threshold	float The threshold used to determine true predictions. Default value: None
overlap_threshold	float If applicable, the bounding box overlap threshold used to determine true predictions. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
IterationPerformance, <xref:msrest.pipeline.ClientRawResponse>	IterationPerformance or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_iterations

Get iterations for the project.

get_iterations(project_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Iteration], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_project

Get a specific project.

get_project(project_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The id of the project to get.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Project, <xref:msrest.pipeline.ClientRawResponse>	Project or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_projects

Get your projects.

get_projects(custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Project], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_tag

Get information about a specific tag.

get_tag(project_id, tag_id, iteration_id=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project this tag belongs to.
tag_id Required	str The tag id.
iteration_id	str The iteration to retrieve this tag from. Optional, defaults to current training set. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Tag, <xref:msrest.pipeline.ClientRawResponse>	Tag or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_tagged_image_count

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.

get_tagged_image_count(project_id, iteration_id=None, tag_ids=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
tag_ids	list[str] A list of tags ids to filter the images to count. Defaults to all tags when null. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
int, <xref:msrest.pipeline.ClientRawResponse>	int or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_tagged_images

Get tagged images for a given project iteration.

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)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
tag_ids	list[str] A list of tags ids to filter the images. Defaults to all tagged images when null. Limited to 20. Default value: None
order_by	str The ordering. Defaults to newest. Possible values include: 'Newest', 'Oldest' Default value: None
take	int Maximum number of images to return. Defaults to 50, limited to 256. Default value: 50
skip	int Number of images to skip before beginning the image batch. Defaults to 0. Default value: 0
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Image], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_tags

Get the tags for a given project and iteration.

get_tags(project_id, iteration_id=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Tag], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_untagged_image_count

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.

get_untagged_image_count(project_id, iteration_id=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
int, <xref:msrest.pipeline.ClientRawResponse>	int or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

get_untagged_images

Get untagged images for a given project iteration.

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

Parameters

Name	Description
project_id Required	str The project id.
iteration_id	str The iteration id. Defaults to workspace. Default value: None
order_by	str The ordering. Defaults to newest. Possible values include: 'Newest', 'Oldest' Default value: None
take	int Maximum number of images to return. Defaults to 50, limited to 256. Default value: 50
skip	int Number of images to skip before beginning the image batch. Defaults to 0. Default value: 0
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[Image], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

import_project

Imports a project.

import_project(token, name=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
token Required	str Token generated from the export project call.
name	str Optional, name of the project to use instead of auto-generated name. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Project, <xref:msrest.pipeline.ClientRawResponse>	Project or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

publish_iteration

Publish a specific iteration.

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

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str The iteration id.
publish_name Required	str The name to give the published iteration.
prediction_id Required	str The id of the prediction resource to publish to.
overwrite	bool Whether to overwrite the published model with the given name (default: false). Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
bool, <xref:msrest.pipeline.ClientRawResponse>	bool or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

query_predictions

Get images that were sent to your prediction endpoint.

query_predictions(project_id, query, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
query Required	PredictionQueryToken Parameters used to query the predictions. Limited to combining 2 tags.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
PredictionQueryResult, <xref:msrest.pipeline.ClientRawResponse>	PredictionQueryResult or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

query_suggested_image_count

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.

query_suggested_image_count(project_id, iteration_id, tag_ids=None, threshold=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str IterationId to use for the suggested tags and regions.
tag_ids	list[str] Existing TagIds in project to get suggested tags count for. Default value: None
threshold	float Confidence threshold to filter suggested tags on. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
dict[str, int], <xref:msrest.pipeline.ClientRawResponse>	dict or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

query_suggested_images

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.

query_suggested_images(project_id, iteration_id, query, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str IterationId to use for the suggested tags and regions.
query Required	SuggestedTagAndRegionQueryToken Contains properties we need to query suggested images.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
SuggestedTagAndRegionQuery, <xref:msrest.pipeline.ClientRawResponse>	SuggestedTagAndRegionQuery or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

quick_test_image

Quick test an image.

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

Parameters

Name	Description
project_id Required	str The project id.
image_data Required	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. Default value: None
store	bool Optional. Specifies whether or not to store the result of this prediction. The default is true, to store. Default value: True
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImagePrediction, <xref:msrest.pipeline.ClientRawResponse>	ImagePrediction or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

quick_test_image_url

Quick test an image url.

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

Parameters

Name	Description
project_id Required	str The project to evaluate against.
url Required	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. Default value: None
store	bool Optional. Specifies whether or not to store the result of this prediction. The default is true, to store. Default value: True
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImagePrediction, <xref:msrest.pipeline.ClientRawResponse>	ImagePrediction or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

suggest_tags_and_regions

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.

suggest_tags_and_regions(project_id, iteration_id, image_ids, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str IterationId to use for tag and region suggestion.
image_ids Required	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 Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
list[SuggestedTagAndRegion], <xref:msrest.pipeline.ClientRawResponse>	list or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

train_project

Queues project for training.

train_project(project_id, training_type=None, reserved_budget_in_hours=0, force_train=False, notification_email_address=None, selected_tags=None, custom_base_model_info=None, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
training_type	str The type of training to use to train the project (default: Regular). Possible values include: 'Regular', 'Advanced' Default value: None
reserved_budget_in_hours	int The number of hours reserved as budget for training (if applicable). Default value: 0
force_train	bool Whether to force train even if dataset and configuration does not change (default: false). Default value: False
notification_email_address	str The email address to send notification to when training finishes (default: null). Default value: None
selected_tags	list[str] List of tags selected for this training session, other tags in the project will be ignored. Default value: None
custom_base_model_info	CustomBaseModelInfo Information of the previously trained iteration which provides the base model for current iteration's training. Default value: None
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Iteration, <xref:msrest.pipeline.ClientRawResponse>	Iteration or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

unpublish_iteration

Unpublish a specific iteration.

unpublish_iteration(project_id, iteration_id, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
iteration_id Required	str The iteration id.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
None, <xref:msrest.pipeline.ClientRawResponse>	None or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

update_image_metadata

Update metadata of images.

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

update_image_metadata(project_id, image_ids, metadata, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
image_ids Required	list[str] The list of image ids to update. Limited to 64.
metadata Required	dict[str, str] The metadata to be updated to the specified images. Limited to 10 key-value pairs per image. The length of key is limited to 128. The length of value is limited to 256.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
ImageMetadataUpdateSummary, <xref:msrest.pipeline.ClientRawResponse>	ImageMetadataUpdateSummary or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

update_iteration

Update a specific iteration.

update_iteration(project_id, iteration_id, name, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str Project id.
iteration_id Required	str Iteration id.
name Required	str Gets or sets the name of the iteration.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Iteration, <xref:msrest.pipeline.ClientRawResponse>	Iteration or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

update_project

Update a specific project.

update_project(project_id, updated_project, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The id of the project to update.
updated_project Required	Project The updated project model.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Project, <xref:msrest.pipeline.ClientRawResponse>	Project or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

update_tag

Update a tag.

update_tag(project_id, tag_id, updated_tag, custom_headers=None, raw=False, **operation_config)

Parameters

Name	Description
project_id Required	str The project id.
tag_id Required	str The id of the target tag.
updated_tag Required	Tag The updated tag model.
custom_headers	dict headers that will be added to the request Default value: None
raw	bool returns the direct response alongside the deserialized response Default value: False
operation_config Required	Operation configuration overrides.

Returns

Type	Description
Tag, <xref:msrest.pipeline.ClientRawResponse>	Tag or ClientRawResponse if raw=true

Exceptions

Type	Description
CustomVisionErrorException

Share via

CustomVisionTrainingClient Class

Constructor

Parameters

Variables

Methods

close

create_image_regions

Parameters

Returns

Exceptions

create_image_tags

Parameters

Returns

Exceptions

create_images_from_data

Parameters

Returns

Exceptions

create_images_from_files

Parameters

Returns

Exceptions

create_images_from_predictions

Parameters

Returns

Exceptions

create_images_from_urls

Parameters

Returns

Exceptions

create_project

Parameters

Returns

Exceptions

create_tag

Parameters

Returns

Exceptions

delete_image_regions

Parameters

Returns

Exceptions

delete_image_tags

Parameters

Returns

Exceptions

delete_images

Parameters

Returns

Exceptions

delete_iteration

Parameters

Returns

Exceptions

delete_prediction

Parameters

Returns

Exceptions

delete_project

Parameters

Returns

Exceptions

delete_tag

Parameters

Returns

Exceptions

export_iteration

Parameters

Returns

Exceptions

export_project

Parameters

Returns

Exceptions

get_artifact

Parameters

Returns

Exceptions

get_domain