FaceOperations Class

FaceOperations operations.

You should not instantiate directly this class, but create a Client instance that will create it for you and attach it as attribute.

Inheritance
builtins.object
FaceOperations

Constructor

FaceOperations(client, config, serializer, deserializer)

Parameters

client
Required

Client for service requests.

config
Required

Configuration of service client.

serializer
Required

An object model serializer.

deserializer
Required

An object model deserializer.

Methods

detect_with_stream

Detect human faces in an image, return face rectangles, and optionally with faceIds, landmarks, and attributes.

  • No image will be stored. Only the extracted face feature will be stored on server. The faceId is an identifier of the face feature and will be used in Face - Identify, Face - Verify, and Face - Find Similar. The stored face feature(s) will expire and be deleted at the time specified by faceIdTimeToLive after the original detection call.
  • Optional parameters include faceId, landmarks, and attributes. Attributes include age, gender, headPose, smile, facialHair, glasses, emotion, hair, makeup, occlusion, accessories, blur, exposure, noise, and mask. Some of the results returned for specific attributes may not be highly accurate.
  • JPEG, PNG, GIF (the first frame), and BMP format are supported. The allowed image file size is from 1KB to 6MB.
  • Up to 100 faces can be returned for an image. Faces are ranked by face rectangle size from large to small.
  • For optimal results when querying Face - Identify, Face - Verify, and Face - Find Similar ('returnFaceId' is true), please use faces that are: frontal, clear, and with a minimum size of 200x200 pixels (100 pixels between eyes).
  • The minimum detectable face size is 36x36 pixels in an image no larger than 1920x1080 pixels. Images with dimensions higher than 1920x1080 pixels will need a proportionally larger minimum face size.
  • Different 'detectionModel' values can be provided. To use and compare different detection models, please refer to How to specify a detection model
  • Different 'recognitionModel' values are provided. If follow-up operations like Verify, Identify, Find Similar are needed, please specify the recognition model with 'recognitionModel' parameter. The default value for 'recognitionModel' is 'recognition_01', if latest model needed, please explicitly specify the model you need in this parameter. Once specified, the detected faceIds will be associated with the specified recognition model. More details, please refer to Specify a recognition model.
detect_with_url

Detect human faces in an image, return face rectangles, and optionally with faceIds, landmarks, and attributes.

  • No image will be stored. Only the extracted face feature will be stored on server. The faceId is an identifier of the face feature and will be used in Face - Identify, Face - Verify, and Face - Find Similar. The stored face feature(s) will expire and be deleted at the time specified by faceIdTimeToLive after the original detection call.
  • Optional parameters include faceId, landmarks, and attributes. Attributes include age, gender, headPose, smile, facialHair, glasses, emotion, hair, makeup, occlusion, accessories, blur, exposure, noise, and mask. Some of the results returned for specific attributes may not be highly accurate.
  • JPEG, PNG, GIF (the first frame), and BMP format are supported. The allowed image file size is from 1KB to 6MB.
  • Up to 100 faces can be returned for an image. Faces are ranked by face rectangle size from large to small.
  • For optimal results when querying Face - Identify, Face - Verify, and Face - Find Similar ('returnFaceId' is true), please use faces that are: frontal, clear, and with a minimum size of 200x200 pixels (100 pixels between eyes).
  • The minimum detectable face size is 36x36 pixels in an image no larger than 1920x1080 pixels. Images with dimensions higher than 1920x1080 pixels will need a proportionally larger minimum face size.
  • Different 'detectionModel' values can be provided. To use and compare different detection models, please refer to How to specify a detection model.
  • Different 'recognitionModel' values are provided. If follow-up operations like Verify, Identify, Find Similar are needed, please specify the recognition model with 'recognitionModel' parameter. The default value for 'recognitionModel' is 'recognition_01', if latest model needed, please explicitly specify the model you need in this parameter. Once specified, the detected faceIds will be associated with the specified recognition model. More details, please refer to Specify a recognition model.
find_similar

Given query face's faceId, to search the similar-looking faces from a faceId array, a face list or a large face list. faceId array contains the faces created by Face - Detect With Url or Face - Detect With Stream, which will expire at the time specified by faceIdTimeToLive after creation. A "faceListId" is created by FaceList - Create containing persistedFaceIds that will not expire. And a "largeFaceListId" is created by LargeFaceList - Create containing persistedFaceIds that will also not expire. Depending on the input the returned similar faces list contains faceIds or persistedFaceIds ranked by similarity.
Find similar has two working modes, "matchPerson" and "matchFace". "matchPerson" is the default mode that it tries to find faces of the same person as possible by using internal same-person thresholds. It is useful to find a known person's other photos. Note that an empty list will be returned if no faces pass the internal thresholds. "matchFace" mode ignores same-person thresholds and returns ranked similar faces anyway, even the similarity is low. It can be used in the cases like searching celebrity-looking faces.
The 'recognitionModel' associated with the query face's faceId should be the same as the 'recognitionModel' used by the target faceId array, face list or large face list. .

group

Divide candidate faces into groups based on face similarity.

  • The output is one or more disjointed face groups and a messyGroup. A face group contains faces that have similar looking, often of the same person. Face groups are ranked by group size, i.e. number of faces. Notice that faces belonging to a same person might be split into several groups in the result.
  • MessyGroup is a special face group containing faces that cannot find any similar counterpart face from original faces. The messyGroup will not appear in the result if all faces found their counterparts.
  • Group API needs at least 2 candidate faces and 1000 at most. We suggest to try Face - Verify when you only have 2 candidate faces.
  • The 'recognitionModel' associated with the query faces' faceIds should be the same. .
identify

1-to-many identification to find the closest matches of the specific query person face from a person group or large person group.
For each face in the faceIds array, Face Identify will compute similarities between the query face and all the faces in the person group (given by personGroupId) or large person group (given by largePersonGroupId), and return candidate person(s) for that face ranked by similarity confidence. The person group/large person group should be trained to make it ready for identification. See more in PersonGroup - Train and LargePersonGroup - Train.
Remarks:

  • The algorithm allows more than one face to be identified independently at the same request, but no more than 10 faces.
  • Each person in the person group/large person group could have more than one face, but no more than 248 faces.
  • Higher face image quality means better identification precision. Please consider high-quality faces: frontal, clear, and face size is 200x200 pixels (100 pixels between eyes) or bigger.
  • Number of candidates returned is restricted by maxNumOfCandidatesReturned and confidenceThreshold. If no person is identified, the returned candidates will be an empty array.
  • Try Face - Find Similar when you need to find similar faces from a face list/large face list instead of a person group/large person group.
  • The 'recognitionModel' associated with the query faces' faceIds should be the same as the 'recognitionModel' used by the target person group or large person group. .
verify_face_to_face

Verify whether two faces belong to a same person or whether one face belongs to a person.
Remarks:

  • Higher face image quality means better identification precision. Please consider high-quality faces: frontal, clear, and face size is 200x200 pixels (100 pixels between eyes) or bigger.
  • For the scenarios that are sensitive to accuracy please make your own judgment.
  • The 'recognitionModel' associated with the query faces' faceIds should be the same as the 'recognitionModel' used by the target face, person group or large person group. .
verify_face_to_person

Verify whether two faces belong to a same person. Compares a face Id with a Person Id.

detect_with_stream

Detect human faces in an image, return face rectangles, and optionally with faceIds, landmarks, and attributes.

  • No image will be stored. Only the extracted face feature will be stored on server. The faceId is an identifier of the face feature and will be used in Face - Identify, Face - Verify, and Face - Find Similar. The stored face feature(s) will expire and be deleted at the time specified by faceIdTimeToLive after the original detection call.
  • Optional parameters include faceId, landmarks, and attributes. Attributes include age, gender, headPose, smile, facialHair, glasses, emotion, hair, makeup, occlusion, accessories, blur, exposure, noise, and mask. Some of the results returned for specific attributes may not be highly accurate.
  • JPEG, PNG, GIF (the first frame), and BMP format are supported. The allowed image file size is from 1KB to 6MB.
  • Up to 100 faces can be returned for an image. Faces are ranked by face rectangle size from large to small.
  • For optimal results when querying Face - Identify, Face - Verify, and Face - Find Similar ('returnFaceId' is true), please use faces that are: frontal, clear, and with a minimum size of 200x200 pixels (100 pixels between eyes).
  • The minimum detectable face size is 36x36 pixels in an image no larger than 1920x1080 pixels. Images with dimensions higher than 1920x1080 pixels will need a proportionally larger minimum face size.
  • Different 'detectionModel' values can be provided. To use and compare different detection models, please refer to How to specify a detection model
  • Different 'recognitionModel' values are provided. If follow-up operations like Verify, Identify, Find Similar are needed, please specify the recognition model with 'recognitionModel' parameter. The default value for 'recognitionModel' is 'recognition_01', if latest model needed, please explicitly specify the model you need in this parameter. Once specified, the detected faceIds will be associated with the specified recognition model. More details, please refer to Specify a recognition model.
detect_with_stream(image, return_face_id=True, return_face_landmarks=False, return_face_attributes=None, recognition_model='recognition_01', return_recognition_model=False, detection_model='detection_01', face_id_time_to_live=86400, custom_headers=None, raw=False, callback=None, **operation_config)

Parameters

image
<xref:Generator>
Required

An image stream.

return_face_id
bool
Required
default value: True

A value indicating whether the operation should return faceIds of detected faces.

return_face_landmarks
bool
Required
default value: False

A value indicating whether the operation should return landmarks of the detected faces.

return_face_attributes
list[str or FaceAttributeType]
Required
default value: None

Analyze and return the one or more specified face attributes in the comma-separated string like "returnFaceAttributes=age,gender". The available attributes depends on the 'detectionModel' specified. 'detection_01' supports age, gender, headPose, smile, facialHair, glasses, emotion, hair, makeup, occlusion, accessories, blur, exposure, and noise. While 'detection_02' does not support any attributes and 'detection_03' only supports mask. Note that each face attribute analysis has additional computational and time cost.

recognition_model
str or RecognitionModel
Required
default value: recognition_01

Name of recognition model. Recognition model is used when the face features are extracted and associated with detected faceIds, (Large)FaceList or (Large)PersonGroup. A recognition model name can be provided when performing Face - Detect or (Large)FaceList - Create or (Large)PersonGroup - Create. The default value is 'recognition_01', if latest model needed, please explicitly specify the model you need. Possible values include: 'recognition_01', 'recognition_02', 'recognition_03', 'recognition_04'

return_recognition_model
bool
Required
default value: False

A value indicating whether the operation should return 'recognitionModel' in response.

detection_model
str or DetectionModel
Required
default value: detection_01

Name of detection model. Detection model is used to detect faces in the submitted image. A detection model name can be provided when performing Face - Detect or (Large)FaceList - Add Face or (Large)PersonGroup - Add Face. The default value is 'detection_01', if another model is needed, please explicitly specify it. Possible values include: 'detection_01', 'detection_02', 'detection_03'

face_id_time_to_live
int
Required
default value: 86400

The number of seconds for the faceId being cached. Supported range from 60 seconds up to 86400 seconds. The default value is 86400 (24 hours).

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

callback
<xref:Callable>[<xref:Bytes>, <xref:response=None>]
Required
default value: 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
Required

Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

Exceptions

detect_with_url

Detect human faces in an image, return face rectangles, and optionally with faceIds, landmarks, and attributes.

  • No image will be stored. Only the extracted face feature will be stored on server. The faceId is an identifier of the face feature and will be used in Face - Identify, Face - Verify, and Face - Find Similar. The stored face feature(s) will expire and be deleted at the time specified by faceIdTimeToLive after the original detection call.
  • Optional parameters include faceId, landmarks, and attributes. Attributes include age, gender, headPose, smile, facialHair, glasses, emotion, hair, makeup, occlusion, accessories, blur, exposure, noise, and mask. Some of the results returned for specific attributes may not be highly accurate.
  • JPEG, PNG, GIF (the first frame), and BMP format are supported. The allowed image file size is from 1KB to 6MB.
  • Up to 100 faces can be returned for an image. Faces are ranked by face rectangle size from large to small.
  • For optimal results when querying Face - Identify, Face - Verify, and Face - Find Similar ('returnFaceId' is true), please use faces that are: frontal, clear, and with a minimum size of 200x200 pixels (100 pixels between eyes).
  • The minimum detectable face size is 36x36 pixels in an image no larger than 1920x1080 pixels. Images with dimensions higher than 1920x1080 pixels will need a proportionally larger minimum face size.
  • Different 'detectionModel' values can be provided. To use and compare different detection models, please refer to How to specify a detection model.
  • Different 'recognitionModel' values are provided. If follow-up operations like Verify, Identify, Find Similar are needed, please specify the recognition model with 'recognitionModel' parameter. The default value for 'recognitionModel' is 'recognition_01', if latest model needed, please explicitly specify the model you need in this parameter. Once specified, the detected faceIds will be associated with the specified recognition model. More details, please refer to Specify a recognition model.
detect_with_url(url, return_face_id=True, return_face_landmarks=False, return_face_attributes=None, recognition_model='recognition_01', return_recognition_model=False, detection_model='detection_01', face_id_time_to_live=86400, custom_headers=None, raw=False, **operation_config)

Parameters

url
str
Required

Publicly reachable URL of an image

return_face_id
bool
Required
default value: True

A value indicating whether the operation should return faceIds of detected faces.

return_face_landmarks
bool
Required
default value: False

A value indicating whether the operation should return landmarks of the detected faces.

return_face_attributes
list[str or FaceAttributeType]
Required
default value: None

Analyze and return the one or more specified face attributes in the comma-separated string like "returnFaceAttributes=age,gender". The available attributes depends on the 'detectionModel' specified. 'detection_01' supports age, gender, headPose, smile, facialHair, glasses, emotion, hair, makeup, occlusion, accessories, blur, exposure, and noise. While 'detection_02' does not support any attributes and 'detection_03' only supports mask. Note that each face attribute analysis has additional computational and time cost.

recognition_model
str or RecognitionModel
Required
default value: recognition_01

Name of recognition model. Recognition model is used when the face features are extracted and associated with detected faceIds, (Large)FaceList or (Large)PersonGroup. A recognition model name can be provided when performing Face - Detect or (Large)FaceList - Create or (Large)PersonGroup - Create. The default value is 'recognition_01', if latest model needed, please explicitly specify the model you need. Possible values include: 'recognition_01', 'recognition_02', 'recognition_03', 'recognition_04'

return_recognition_model
bool
Required
default value: False

A value indicating whether the operation should return 'recognitionModel' in response.

detection_model
str or DetectionModel
Required
default value: detection_01

Name of detection model. Detection model is used to detect faces in the submitted image. A detection model name can be provided when performing Face - Detect or (Large)FaceList - Add Face or (Large)PersonGroup - Add Face. The default value is 'detection_01', if another model is needed, please explicitly specify it. Possible values include: 'detection_01', 'detection_02', 'detection_03'

face_id_time_to_live
int
Required
default value: 86400

The number of seconds for the faceId being cached. Supported range from 60 seconds up to 86400 seconds. The default value is 86400 (24 hours).

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

operation_config
Required

Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

Exceptions

find_similar

Given query face's faceId, to search the similar-looking faces from a faceId array, a face list or a large face list. faceId array contains the faces created by Face - Detect With Url or Face - Detect With Stream, which will expire at the time specified by faceIdTimeToLive after creation. A "faceListId" is created by FaceList - Create containing persistedFaceIds that will not expire. And a "largeFaceListId" is created by LargeFaceList - Create containing persistedFaceIds that will also not expire. Depending on the input the returned similar faces list contains faceIds or persistedFaceIds ranked by similarity.
Find similar has two working modes, "matchPerson" and "matchFace". "matchPerson" is the default mode that it tries to find faces of the same person as possible by using internal same-person thresholds. It is useful to find a known person's other photos. Note that an empty list will be returned if no faces pass the internal thresholds. "matchFace" mode ignores same-person thresholds and returns ranked similar faces anyway, even the similarity is low. It can be used in the cases like searching celebrity-looking faces.
The 'recognitionModel' associated with the query face's faceId should be the same as the 'recognitionModel' used by the target faceId array, face list or large face list. .

find_similar(face_id, face_list_id=None, large_face_list_id=None, face_ids=None, max_num_of_candidates_returned=20, mode='matchPerson', custom_headers=None, raw=False, **operation_config)

Parameters

face_id
str
Required

FaceId of the query face. User needs to call Face - Detect first to get a valid faceId. Note that this faceId is not persisted and will expire at the time specified by faceIdTimeToLive after the detection call

face_list_id
str
Required
default value: None

An existing user-specified unique candidate face list, created in Face List - Create a Face List. Face list contains a set of persistedFaceIds which are persisted and will never expire. Parameter faceListId, largeFaceListId and faceIds should not be provided at the same time.

large_face_list_id
str
Required
default value: None

An existing user-specified unique candidate large face list, created in LargeFaceList - Create. Large face list contains a set of persistedFaceIds which are persisted and will never expire. Parameter faceListId, largeFaceListId and faceIds should not be provided at the same time.

face_ids
list[str]
Required
default value: None

An array of candidate faceIds. All of them are created by Face - Detect and the faceIds will expire at the time specified by faceIdTimeToLive after the detection call. The number of faceIds is limited to 1000. Parameter faceListId, largeFaceListId and faceIds should not be provided at the same time.

max_num_of_candidates_returned
int
Required
default value: 20

The number of top similar faces returned. The valid range is [1, 1000].

mode
str or FindSimilarMatchMode
Required
default value: matchPerson

Similar face searching mode. It can be "matchPerson" or "matchFace". Possible values include: 'matchPerson', 'matchFace'

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

operation_config
Required

Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

Exceptions

group

Divide candidate faces into groups based on face similarity.

  • The output is one or more disjointed face groups and a messyGroup. A face group contains faces that have similar looking, often of the same person. Face groups are ranked by group size, i.e. number of faces. Notice that faces belonging to a same person might be split into several groups in the result.
  • MessyGroup is a special face group containing faces that cannot find any similar counterpart face from original faces. The messyGroup will not appear in the result if all faces found their counterparts.
  • Group API needs at least 2 candidate faces and 1000 at most. We suggest to try Face - Verify when you only have 2 candidate faces.
  • The 'recognitionModel' associated with the query faces' faceIds should be the same. .
group(face_ids, custom_headers=None, raw=False, **operation_config)

Parameters

face_ids
list[str]
Required

Array of candidate faceId created by Face - Detect. The maximum is 1000 faces

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

operation_config
Required

Operation configuration overrides.

Returns

GroupResult or ClientRawResponse if raw=true

Return type

Exceptions

identify

1-to-many identification to find the closest matches of the specific query person face from a person group or large person group.
For each face in the faceIds array, Face Identify will compute similarities between the query face and all the faces in the person group (given by personGroupId) or large person group (given by largePersonGroupId), and return candidate person(s) for that face ranked by similarity confidence. The person group/large person group should be trained to make it ready for identification. See more in PersonGroup - Train and LargePersonGroup - Train.
Remarks:

  • The algorithm allows more than one face to be identified independently at the same request, but no more than 10 faces.
  • Each person in the person group/large person group could have more than one face, but no more than 248 faces.
  • Higher face image quality means better identification precision. Please consider high-quality faces: frontal, clear, and face size is 200x200 pixels (100 pixels between eyes) or bigger.
  • Number of candidates returned is restricted by maxNumOfCandidatesReturned and confidenceThreshold. If no person is identified, the returned candidates will be an empty array.
  • Try Face - Find Similar when you need to find similar faces from a face list/large face list instead of a person group/large person group.
  • The 'recognitionModel' associated with the query faces' faceIds should be the same as the 'recognitionModel' used by the target person group or large person group. .
identify(face_ids, person_group_id=None, large_person_group_id=None, max_num_of_candidates_returned=1, confidence_threshold=None, custom_headers=None, raw=False, **operation_config)

Parameters

face_ids
list[str]
Required

Array of query faces faceIds, created by the Face - Detect. Each of the faces are identified independently. The valid number of faceIds is between [1, 10].

person_group_id
str
Required
default value: None

PersonGroupId of the target person group, created by PersonGroup - Create. Parameter personGroupId and largePersonGroupId should not be provided at the same time.

large_person_group_id
str
Required
default value: None

LargePersonGroupId of the target large person group, created by LargePersonGroup - Create. Parameter personGroupId and largePersonGroupId should not be provided at the same time.

max_num_of_candidates_returned
int
Required
default value: 1

The range of maxNumOfCandidatesReturned is between 1 and 5 (default is 1).

confidence_threshold
float
Required
default value: None

Confidence threshold of identification, used to judge whether one face belong to one person. The range of confidenceThreshold is [0, 1] (default specified by algorithm).

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

operation_config
Required

Operation configuration overrides.

Returns

list or ClientRawResponse if raw=true

Return type

Exceptions

verify_face_to_face

Verify whether two faces belong to a same person or whether one face belongs to a person.
Remarks:

  • Higher face image quality means better identification precision. Please consider high-quality faces: frontal, clear, and face size is 200x200 pixels (100 pixels between eyes) or bigger.
  • For the scenarios that are sensitive to accuracy please make your own judgment.
  • The 'recognitionModel' associated with the query faces' faceIds should be the same as the 'recognitionModel' used by the target face, person group or large person group. .
verify_face_to_face(face_id1, face_id2, custom_headers=None, raw=False, **operation_config)

Parameters

face_id1
str
Required

FaceId of the first face, comes from Face - Detect

face_id2
str
Required

FaceId of the second face, comes from Face - Detect

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

operation_config
Required

Operation configuration overrides.

Returns

VerifyResult or ClientRawResponse if raw=true

Return type

Exceptions

verify_face_to_person

Verify whether two faces belong to a same person. Compares a face Id with a Person Id.

verify_face_to_person(face_id, person_id, person_group_id=None, large_person_group_id=None, custom_headers=None, raw=False, **operation_config)

Parameters

face_id
str
Required

FaceId of the face, comes from Face - Detect

person_id
str
Required

Specify a certain person in a person group or a large person group. personId is created in PersonGroup Person - Create or LargePersonGroup Person - Create.

person_group_id
str
Required
default value: None

Using existing personGroupId and personId for fast loading a specified person. personGroupId is created in PersonGroup - Create. Parameter personGroupId and largePersonGroupId should not be provided at the same time.

large_person_group_id
str
Required
default value: None

Using existing largePersonGroupId and personId for fast loading a specified person. largePersonGroupId is created in LargePersonGroup - Create. Parameter personGroupId and largePersonGroupId should not be provided at the same time.

custom_headers
dict
Required
default value: None

headers that will be added to the request

raw
bool
Required
default value: False

returns the direct response alongside the deserialized response

operation_config
Required

Operation configuration overrides.

Returns

VerifyResult or ClientRawResponse if raw=true

Return type

Exceptions

Attributes

models

models = <module 'azure.cognitiveservices.vision.face.models' from 'c:\\hostedtoolcache\\windows\\python\\3.6.8\\x64\\lib\\site-packages\\azure\\cognitiveservices\\vision\\face\\models\\__init__.py'>