CustomVisionPredictionClient Class

Inheritance

azure.cognitiveservices.vision.customvision.prediction.operations._custom_vision_prediction_client_operations.CustomVisionPredictionClientOperationsMixin

CustomVisionPredictionClient

SDKClient

CustomVisionPredictionClient

Constructor

CustomVisionPredictionClient(endpoint, credentials)

Parameters

endpoint

str

Required

Supported Cognitive Services endpoints.

credentials

None

Required

Subscription credentials which uniquely identify client subscription.

Variables

config

CustomVisionPredictionClientConfiguration

Configuration for client.

Methods

classify_image	Classify an image and saves the result.
classify_image_url	Classify an image url and saves the result.
classify_image_url_with_no_store	Classify an image url without saving the result.
classify_image_with_no_store	Classify an image without saving the result.
close	Close the client if keep_alive is True.
detect_image	Detect objects in an image and saves the result.
detect_image_url	Detect objects in an image url and saves the result.
detect_image_url_with_no_store	Detect objects in an image url without saving the result.
detect_image_with_no_store	Detect objects in an image without saving the result.

classify_image

Classify an image and saves the result.

classify_image(project_id, published_name, image_data, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

image_data

Generator

Required

Binary image data. Supported formats are JPEG, GIF, PNG, and BMP. Supports images up to 4MB.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

classify_image_url

Classify an image url and saves the result.

classify_image_url(project_id, published_name, url, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

url

str

Required

Url of the image.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

classify_image_url_with_no_store

Classify an image url without saving the result.

classify_image_url_with_no_store(project_id, published_name, url, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

url

str

Required

Url of the image.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

classify_image_with_no_store

Classify an image without saving the result.

classify_image_with_no_store(project_id, published_name, image_data, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

image_data

Generator

Required

Binary image data. Supported formats are JPEG, GIF, PNG, and BMP. Supports images up to 4MB.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

close

Close the client if keep_alive is True.

close() -> None

Exceptions

CustomVisionErrorException

detect_image

Detect objects in an image and saves the result.

detect_image(project_id, published_name, image_data, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

image_data

Generator

Required

Binary image data. Supported formats are JPEG, GIF, PNG, and BMP. Supports images up to 4MB.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

detect_image_url

Detect objects in an image url and saves the result.

detect_image_url(project_id, published_name, url, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

url

str

Required

Url of the image.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

detect_image_url_with_no_store

Detect objects in an image url without saving the result.

detect_image_url_with_no_store(project_id, published_name, url, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

url

str

Required

Url of the image.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException

detect_image_with_no_store

Detect objects in an image without saving the result.

detect_image_with_no_store(project_id, published_name, image_data, application=None, custom_headers=None, raw=False, **operation_config)

Parameters

project_id

str

Required

The project id.

published_name

str

Required

Specifies the name of the model to evaluate against.

image_data

Generator

Required

Binary image data. Supported formats are JPEG, GIF, PNG, and BMP. Supports images up to 4MB.

application

str

default value: None

Optional. Specifies the name of application using the endpoint.

custom_headers

dict

default value: None

headers that will be added to the request

raw

bool

default value: False

returns the direct response alongside the deserialized response

operation_config

Required

Operation configuration overrides.

Returns

ImagePrediction or ClientRawResponse if raw=true

Return type

ImagePrediction,

ClientRawResponse

Exceptions

CustomVisionErrorException