Face attribute information
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
Type | Integer | Attribute value |
Probability | Float | Probability of recognizing Type , which indicates the probability of correct recognition. Value range: [0,1]. |
Most matching candidate recognized
Used by actions: SearchFaces, SearchFacesReturnsByGroup, SearchPersons, SearchPersonsReturnsByGroup.
Name | Type | Description |
---|---|---|
PersonId | String | Person ID |
FaceId | String | Face ID, which is valid only when returned by the SearchFaces or SearchFacesReturnsByGroup API. User search APIs use facial feature fusion to search for users, for which this field is meaningless. |
Score | Float | Match score of candidate. In a face base library containing 10,000 faces, the 1%, 0.1%, and 0.01% FARs correspond to scores of 70, 80, and 90, respectively; In a face base library containing 100,000 faces, the 1%, 0.1%, and 0.01% FARs correspond to scores of 80, 90, and 100, respectively; In a face base library containing 300,000 faces, the 1% and 0.1% FARs correspond to scores of 85 and 95, respectively. Generally, the score of 80 is suitable for most scenarios. We recommend choosing an appropriate score based on the actual situation, preferably no more than 90. |
PersonName | String | Person name Note: this field may return null, indicating that no valid values can be obtained. |
Gender | Integer | Person gender Note: this field may return null, indicating that no valid values can be obtained. |
PersonGroupInfos | Array of PersonGroupInfo | List of groups containing this person and their description fields Note: this field may return null, indicating that no valid values can be obtained. |
Eye information
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
Glass | AttributeItem | Whether glasses are worn. The Type values of the AttributeItem include: 0: no glasses; 1: general glasses; 2: sunglasses. |
EyeOpen | AttributeItem | Whether the eyes are open. The Type values of the AttributeItem include: 0: open; 1: closed. |
EyelidType | AttributeItem | Whether the person has double eyelids. The Type values of the AttributeItem include: 0: no; 1: yes. |
EyeSize | AttributeItem | Eye size. The Type values of the AttributeItem include: 0: small eyes; 1: general eyes; 2: big eyes. |
Eyebrow information
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
EyebrowDensity | AttributeItem | Eyebrow thickness. The Type values of the AttributeItem include: 0: light; 1: thick. |
EyebrowCurve | AttributeItem | Eyebrow curve. The Type values of the AttributeItem include: 0: flat; 1: curved. |
EyebrowLength | AttributeItem | Eyebrow length. The Type values of the AttributeItem include: 0: short; 1: long. |
Face attributes, including gender, age, expression,
beauty, glass, mask, hair, and pose (pitch, roll, yaw). Valid information will be returned only if NeedFaceAttributes
is set to 1. The face attribute information of up to 5 largest faces in the image will be returned, and FaceAttributesInfo
of the 6th and rest faces is meaningless.
Used by actions: DetectFace.
Name | Type | Description |
---|---|---|
Gender | Integer | Gender. The gender is female for the value range [0,49] and male for the value range [50,100]. The closer the value to 0 or 100, the higher the confidence. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Age | Integer | Age. Value range: [0,100]. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Expression | Integer | Expression. Value range: [0 (normal)–50 (smile)–100 (laugh)]. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Glass | Boolean | Whether glasses are present. Valid values: true, false. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Pitch | Integer | Vertical offset in degrees. Value range: [-30,30]. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. We recommend selecting images in the [-10,10] range for adding faces. |
Yaw | Integer | Horizontal offset in degrees. Value range: [-30,30]. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. We recommend selecting images in the [-10,10] range for adding faces. |
Roll | Integer | Horizontal rotation in degrees. Value range: [-180,180]. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. We recommend selecting images in the [-20,20] range for adding faces. |
Beauty | Integer | Beauty. Value range: [0,100]. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Hat | Boolean | Whether hat is present. Valid values: true, false. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless.Note: this field may return null, indicating that no valid values can be obtained. |
Mask | Boolean | Whether mask is present. Valid values: true, false. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless.Note: this field may return null, indicating that no valid values can be obtained. |
Hair | FaceHairAttributesInfo | Hair information, including length, bang, and color. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless.Note: this field may return null, indicating that no valid values can be obtained. |
EyeOpen | Boolean | Whether the eyes are open. Valid values: true, false. As long as there is more than one eye closed, false will be returned. If NeedFaceAttributes is not 1 or more than 5 faces are detected, this parameter will still be returned but meaningless.Note: this field may return null, indicating that no valid values can be obtained. |
Face attribute information. According to the types specified in FaceAttributesType
, the following face attributes will be returned: Age, Beauty,
Emotion, Eye, Eyebrow, Gender,
Hair, Hat, Headpose, Mask, Mouth, Moustache,
Nose, Shape, Skin, Smile, etc.
If no types are specified in FaceAttributesType
, the details returned by FaceDetaiAttributesInfo
will be meaningless.
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
Age | Integer | Age. Value range: [0,65], where 65 indicates 65 years old or above. If FaceAttributesType does not include Age or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Beauty | Integer | Beauty score. Value range: [0,100]. If FaceAttributesType does not include Beauty or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Emotion | AttributeItem | Emotion, including relaxed, happy, surprised, angry, sad, disgusted, and scared. The Type values of the AttributeItem include: 0: relaxed; 1: happy; 2: surprised; 3: angry; 4: sad; 5: disgusted; 6: scared.If FaceAttributesType does not include Emotion or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Eye | Eye | Eye information, including whether glasses are worn, whether eyes are closed, whether the person has double eyelids, and the eye size. If FaceAttributesType does not include Eye or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Eyebrow | Eyebrow | Eyebrow information, including whether the eyebrows are thick, curved, or long. If FaceAttributesType does not include Eyebrow or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Gender | AttributeItem | Gender information. The Type values of the AttributeItem include: 0: male; 1: female. If FaceAttributesType does not include Gender or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Hair | Hair | Hair information, including length, bang, and color. If FaceAttributesType does not include Hair or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Hat | Hat | Hat information, including whether a hat is worn, hat style, and hat color. If FaceAttributesType does not include Hat or more than 5 faces are detected, this parameter will still be returned but meaningless. |
HeadPose | HeadPose | Pose information, including the face pitch, yaw, and roll. If FaceAttributesType does not include Headpose or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Mask | AttributeItem | Mask information. The Type values of the AttributeItem include: 0: no mask; 1: the mask is worn and does not cover the face; 2: the mask is worn and covers the chin; 3: the mask is worn and covers the mouth; 4: the mask is worn properly.If FaceAttributesType does not include Mask or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Mouth | Mouth | Mouth information, including whether the mouth is open and the lip thickness. If FaceAttributesType does not include Mouth or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Moustache | AttributeItem | Beard information. The Type values of the AttributeItem include: 0: no beard; 1: beard detected. If FaceAttributesType does not include Moustache or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Nose | AttributeItem | Nose information. The Type values of the AttributeItem include: 0: upturned nose; 1: aquiline nose; 2: general nose; 3: bulbous nose.If FaceAttributesType does not include Nose or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Shape | AttributeItem | Face shape information. The Type values of the AttributeItem include: 0: square; 1: triangular; 2: oval; 3: heart-shaped; 4: round.If FaceAttributesType does not include Shape or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Skin | AttributeItem | Skin color information. The Type values of the AttributeItem include: 0: yellow; 1: brown; 2: black; 3: white.If FaceAttributesType does not include Skin or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Smile | Integer | Smile level. Value range: [0,100]. If FaceAttributesType does not include Smile or more than 5 faces are detected, this parameter will still be returned but meaningless. |
Face information list.
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
FaceRect | FaceRect | Position of the detected face frame. |
FaceDetailAttributesInfo | FaceDetailAttributesInfo | Face attribute information. According to the types specified in FaceAttributesType , the following face attributes will be returned: age (Age), beauty score (Beauty), emotion (Emotion), eye information (Eye), eyebrow information (Eyebrow), gender (Gender), hair information (Hair), hat information (Hat), pose (Headpose), mask information (Mask), mouth information (Mouse), beard information (Moustache), nose information (Nose), face shape (Shape), skin color (Skin), and smile information (Smile), etc. If no types are specified in FaceAttributesType , the detailed items returned by FaceDetaiAttributesInfo will be meaningless. |
Hair information in face attributes.
Used by actions: DetectFace.
Name | Type | Description |
---|---|---|
Length | Integer | 0: shaved head, 1: short hair, 2: medium hair, 3: long hair, 4: braid Note: this field may return null, indicating that no valid values can be obtained. |
Bang | Integer | 0: with bangs, 1: no bangs Note: this field may return null, indicating that no valid values can be obtained. |
Color | Integer | 0: black, 1: golden, 2: brown, 3: gray Note: this field may return null, indicating that no valid values can be obtained. |
Face information list.
Used by actions: DetectFace.
Name | Type | Description |
---|---|---|
X | Integer | Horizontal coordinate of the top-left vertex of the face frame. The face frame encompasses the facial features and is extended accordingly. If it is larger than the image, the coordinates will be negative. If you want to capture a complete face, you can set the negative coordinates to 0 if the completeness score meets the requirement. |
Y | Integer | Vertical coordinate of the top-left vertex of the face frame. The face frame encompasses the facial features and is extended accordingly. If it is larger than the image, the coordinates will be negative. If you want to capture a complete face, you can set the negative coordinates to 0 if the completeness score meets the requirement. |
Width | Integer | Face frame width. |
Height | Integer | Face frame height. |
FaceAttributesInfo | FaceAttributesInfo | Face attributes, including gender, age, expression, beauty, glass, mask, hair, and pose (pitch, roll, yaw). Valid information will be returned only if NeedFaceAttributes is set to 1. |
FaceQualityInfo | FaceQualityInfo | Face quality information, including score, sharpness, brightness, and completeness. Valid information will be returned only if NeedFaceDetection is set to 1.Note: this field may return null, indicating that no valid values can be obtained. |
Completeness of facial features, which assesses the completeness of the eyebrows, eyes, nose, cheeks, mouth, and chin.
Used by actions: DetectFace.
Name | Type | Description |
---|---|---|
Eyebrow | Integer | Eyebrow completeness. Value range: [0,100]. The higher the score, the higher the completeness. Reference range: [0,80], which means incomplete. Note: this field may return null, indicating that no valid values can be obtained. |
Eye | Integer | Eye completeness. Value range: [0,100]. The higher the score, the higher the completeness. Reference range: [0,80], which means incomplete. Note: this field may return null, indicating that no valid values can be obtained. |
Nose | Integer | Nose completeness. Value range: [0,100]. The higher the score, the higher the completeness. Reference range: [0,60], which means incomplete. Note: this field may return null, indicating that no valid values can be obtained. |
Cheek | Integer | Cheek completeness. Value range: [0,100]. The higher the score, the higher the completeness. Reference range: [0,70], which means incomplete. Note: this field may return null, indicating that no valid values can be obtained. |
Mouth | Integer | Mouth completeness. Value range: [0,100]. The higher the score, the higher the completeness. Reference range: [0,50], which means incomplete. Note: this field may return null, indicating that no valid values can be obtained. |
Chin | Integer | Chin completeness. Value range: [0,100]. The higher the score, the higher the completeness. Reference range: [0,70], which means incomplete. Note: this field may return null, indicating that no valid values can be obtained. |
Face quality information, including score, sharpness, brightness, and completeness. Valid information will be returned only if NeedFaceDetection
is set to 1.
Used by actions: DetectFace.
Name | Type | Description |
---|---|---|
Score | Integer | Quality score. Value range: [0,100]. It comprehensively evaluates whether the image quality is suitable for face recognition; the higher the score, the higher the quality. In normal cases, you only need to use Score as the overall quality standard score. Specific item scores such as Sharpness , Brightness , Completeness are for reference only.Reference range: [0,40]: poor; [40,60]: fine; [60,80]: good; [80,100]: excellent. We recommend selecting images with a score above 70 for adding faces. Note: this field may return null, indicating that no valid values can be obtained. |
Sharpness | Integer | Sharpness. Value range: [0,100]. It evaluates the sharpness of the image. The higher the score, the sharper the image. Reference range: [0,40]: very blurry; [40,60]: blurry; [60,80]: fine; [80,100]: sharp. We recommend selecting images with a score above 80 for adding faces. Note: this field may return null, indicating that no valid values can be obtained. |
Brightness | Integer | Brightness. Value range: [0,100]. The brighter the image, the higher the score. Reference range: [0,30]: dark; [30,70]: normal; [70,100]: bright. We recommend selecting images in the [30,70] range for adding faces. Note: this field may return null, indicating that no valid values can be obtained. |
Completeness | FaceQualityCompleteness | Completeness of facial features, which assesses the completeness of the eyebrows, eyes, nose, cheeks, mouth, and chin. Note: this field may return null, indicating that no valid values can be obtained. |
Position of detected face frame
Used by actions: CreateFace, CreatePerson, DetectFaceAttributes, SearchFaces, SearchFacesReturnsByGroup, SearchPersons, SearchPersonsReturnsByGroup.
Name | Type | Description |
---|---|---|
X | Integer | Horizontal coordinate of the top-left vertex of face frame. The face frame encompasses the facial features and is extended accordingly. If it is larger than the image, the coordinates will be negative. If you want to capture a complete face, you can set the negative coordinates to 0 if the completeness score meets the requirement. |
Y | Integer | Vertical coordinate of the top-left vertex of face frame. The face frame encompasses the facial features and is extended accordingly. If it is larger than the image, the coordinates will be negative. If you want to capture a complete face, you can set the negative coordinates to 0 if the completeness score meets the requirement. |
Width | Integer | Face width |
Height | Integer | Face height |
Specific information of facial feature localization (facial keypoints).
Used by actions: AnalyzeFace.
Name | Type | Description |
---|---|---|
FaceProfile | Array of Point | 21 points that describe the face contour. |
LeftEye | Array of Point | 8 points that describe the left eye. |
RightEye | Array of Point | 8 points that describe the right eye. |
LeftEyeBrow | Array of Point | 8 points that describe the left eyebrow. |
RightEyeBrow | Array of Point | 8 points that describe the right eyebrow. |
Mouth | Array of Point | 22 points that describe the mouth. |
Nose | Array of Point | 13 points that describe the nose. |
LeftPupil | Array of Point | 1 point that describes the left pupil. |
RightPupil | Array of Point | 1 point that describes the right pupil. |
Recognition result items by group
Used by actions: SearchFacesReturnsByGroup, SearchPersonsReturnsByGroup.
Name | Type | Description |
---|---|---|
GroupId | String | Group ID. |
Candidates | Array of Candidate | Most matching candidate recognized |
Custom description field of the group to be modified, which is a key-value
pair.
Used by actions: ModifyGroup.
Name | Type | Required | Description |
---|---|---|---|
GroupExDescriptionIndex | Integer | Yes | Custom group description field index, whose value starts from 0. Note: this field may return null, indicating that no valid values can be obtained. |
GroupExDescription | String | Yes | Content of the custom group description field to be updated |
Returned group information
Used by actions: GetGroupList.
Name | Type | Description |
---|---|---|
GroupName | String | Group name |
GroupId | String | Group ID |
GroupExDescriptions | Array of String | Custom group description field Note: this field may return null, indicating that no valid values can be obtained. |
Tag | String | Group remarks Note: this field may return null, indicating that no valid values can be obtained. |
FaceModelVersion | String | Algorithm model version used for face recognition. Note: this field may return null, indicating that no valid values can be obtained. |
CreationTimestamp | Integer | Group creation time and date (CreationTimestamp ), whose value is the number of milliseconds between the UNIX epoch time and the group creation time. The UNIX epoch time is 00:00:00, Thursday, January 1, 1970, Coordinated Universal Time (UTC). For more information, please see the UNIX time document. Note: this field may return null, indicating that no valid values can be obtained. |
Hair information
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
Length | AttributeItem | Hair length information. The Type values of the AttributeItem include: 0: bald, 1: short hair, 2: medium hair, 3: long hair, 4: braid. |
Bang | AttributeItem | Bang information. The Type values of the AttributeItem include: 0: no bang; 1: bang detected. |
Color | AttributeItem | Hair color information. The Type values of the AttributeItem include: 0: black; 1: golden; 2: brown; 3: gray. |
Hat information
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
Style | AttributeItem | Hat wearing status information. The Type values of the AttributeItem include: 0: no hat; 1: general hat; 2: helmet; 3: security guard hat. |
Color | AttributeItem | Hat color. The Type values of the AttributeItem include: 0: no hat; 1: red; 2: yellow; 3: blue; 4: black; 5: gray; 6: mixed colors. |
Pose information.
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
Pitch | Integer | Pitch. Value range: [-30,30]. |
Yaw | Integer | Yaw. Value range: [-30,30]. |
Roll | Integer | Roll. Value range: [-180,180]. |
Mouth information
Used by actions: DetectFaceAttributes.
Name | Type | Description |
---|---|---|
MouthOpen | AttributeItem | Whether the mouth is open. The Type values of the AttributeItem include: 0: closed; 1: open. |
Custom description field of the person to be modified, which is a key-value
pair.
Used by actions: CreatePerson, ModifyPersonGroupInfo.
Name | Type | Required | Description |
---|---|---|---|
PersonExDescriptionIndex | Integer | Yes | Person description field index, whose value starts from 0. Note: this field may return null, indicating that no valid values can be obtained. |
PersonExDescription | String | Yes | Content of the person description field to be updated |
List of groups containing this person and their description fields
Used by actions: GetPersonGroupInfo, SearchFaces, SearchFacesReturnsByGroup, SearchPersons, SearchPersonsReturnsByGroup.
Name | Type | Description |
---|---|---|
GroupId | String | ID of the group that contains this person |
PersonExDescriptions | Array of String | Content of person description field |
Returned person information
Used by actions: GetPersonList.
Name | Type | Description |
---|---|---|
PersonName | String | Person name |
PersonId | String | Person ID |
Gender | Integer | Person gender |
PersonExDescriptions | Array of String | Content of person description field |
FaceIds | Array of String | List of contained face images |
CreationTimestamp | Integer | Person creation time, measured in the number of milliseconds elapsed since the Unix epoch The Unix epoch is 00:00:00, Thursday, January 1, 1970, Coordinated Universal Time (UTC). For more information, please see the Unix time document. |
Coordinates
Used by actions: AnalyzeFace.
Name | Type | Description |
---|---|---|
X | Integer | X coordinate |
Y | Integer | Y coordinate |
Face recognition result
Used by actions: SearchFaces, SearchPersons.
Name | Type | Description |
---|---|---|
Candidates | Array of Candidate | Most matching candidate recognized |
FaceRect | FaceRect | Position of detected face frame |
RetCode | Integer | The status return code for the face image detected. Valid values: 0 - normal; -1601 - the image does not meet the quality requirements, in which case Candidate is empty; -1604 - the face similarity is not higher than FaceMatchThreshold . |
Recognition result.
Used by actions: SearchFacesReturnsByGroup, SearchPersonsReturnsByGroup.
Name | Type | Description |
---|---|---|
FaceRect | FaceRect | Position of detected face frame |
GroupCandidates | Array of GroupCandidate | Recognition result. |
RetCode | Integer | Status return code of detected face image. 0: normal. -1601: the image quality control requirement is not met; in this case, Candidate is empty. |
Was this page helpful?