View Source Evision.FaceRecognizerSF (Evision v0.2.9)

Summary

Types

t()

Type that represents an FaceRecognizerSF struct.

Functions

alignCrop(named_args)

alignCrop(self, src_img, face_box)

Aligning image to put face on the standard position

alignCrop(self, src_img, face_box, opts)

Aligning image to put face on the standard position

create(named_args)

create(model, config)

Creates an instance of this class with given parameters

create(model, config, opts)

Creates an instance of this class with given parameters

feature(named_args)

feature(self, aligned_img)

Extracting face feature from aligned image

feature(self, aligned_img, opts)

Extracting face feature from aligned image

match(named_args)

match(self, face_feature1, face_feature2)

Calculating the distance between two face features

match(self, face_feature1, face_feature2, opts)

Calculating the distance between two face features

Types

t()

@type t() :: %Evision.FaceRecognizerSF{ref: reference()}

Type that represents an FaceRecognizerSF struct.

ref. reference()
The underlying erlang resource variable.

Functions

alignCrop(named_args)

@spec alignCrop(Keyword.t()) :: any() | {:error, String.t()}

alignCrop(self, src_img, face_box)

@spec alignCrop(t(), Evision.Mat.maybe_mat_in(), Evision.Mat.maybe_mat_in()) ::
  Evision.Mat.t() | {:error, String.t()}

Aligning image to put face on the standard position

Positional Arguments

self: Evision.FaceRecognizerSF.t()
src_img: Evision.Mat.
input image
face_box: Evision.Mat.
the detection result used for indicate face in input image

Return

aligned_img: Evision.Mat.t().
output aligned image

Python prototype (for reference only):

alignCrop(src_img, face_box[, aligned_img]) -> aligned_img

alignCrop(self, src_img, face_box, opts)

@spec alignCrop(
  t(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in(),
  [{atom(), term()}, ...] | nil
) :: Evision.Mat.t() | {:error, String.t()}

Aligning image to put face on the standard position

Positional Arguments

self: Evision.FaceRecognizerSF.t()
src_img: Evision.Mat.
input image
face_box: Evision.Mat.
the detection result used for indicate face in input image

Return

aligned_img: Evision.Mat.t().
output aligned image

Python prototype (for reference only):

alignCrop(src_img, face_box[, aligned_img]) -> aligned_img

create(named_args)

@spec create(Keyword.t()) :: any() | {:error, String.t()}

create(model, config)

@spec create(binary(), binary()) :: t() | {:error, String.t()}

Creates an instance of this class with given parameters

Positional Arguments

model: String.
the path of the onnx model used for face recognition
config: String.
the path to the config file for compability, which is not requested for ONNX models

Keyword Arguments

backend_id: integer().
the id of backend
target_id: integer().
the id of target device

Return

retval: Evision.FaceRecognizerSF.t()

Python prototype (for reference only):

create(model, config[, backend_id[, target_id]]) -> retval

create(model, config, opts)

@spec create(binary(), binary(), [backend_id: term(), target_id: term()] | nil) ::
  t() | {:error, String.t()}

Creates an instance of this class with given parameters

Positional Arguments

model: String.
the path of the onnx model used for face recognition
config: String.
the path to the config file for compability, which is not requested for ONNX models

Keyword Arguments

backend_id: integer().
the id of backend
target_id: integer().
the id of target device

Return

retval: Evision.FaceRecognizerSF.t()

Python prototype (for reference only):

create(model, config[, backend_id[, target_id]]) -> retval

feature(named_args)

@spec feature(Keyword.t()) :: any() | {:error, String.t()}

feature(self, aligned_img)

@spec feature(t(), Evision.Mat.maybe_mat_in()) ::
  Evision.Mat.t() | {:error, String.t()}

Extracting face feature from aligned image

Positional Arguments

self: Evision.FaceRecognizerSF.t()
aligned_img: Evision.Mat.
input aligned image

Return

face_feature: Evision.Mat.t().
output face feature

Python prototype (for reference only):

feature(aligned_img[, face_feature]) -> face_feature

feature(self, aligned_img, opts)

@spec feature(t(), Evision.Mat.maybe_mat_in(), [{atom(), term()}, ...] | nil) ::
  Evision.Mat.t() | {:error, String.t()}

Extracting face feature from aligned image

Positional Arguments

self: Evision.FaceRecognizerSF.t()
aligned_img: Evision.Mat.
input aligned image

Return

face_feature: Evision.Mat.t().
output face feature

Python prototype (for reference only):

feature(aligned_img[, face_feature]) -> face_feature

match(named_args)

@spec match(Keyword.t()) :: any() | {:error, String.t()}

match(self, face_feature1, face_feature2)

@spec match(t(), Evision.Mat.maybe_mat_in(), Evision.Mat.maybe_mat_in()) ::
  number() | {:error, String.t()}

Calculating the distance between two face features

Positional Arguments

self: Evision.FaceRecognizerSF.t()
face_feature1: Evision.Mat.
the first input feature
face_feature2: Evision.Mat.
the second input feature of the same size and the same type as face_feature1

Keyword Arguments

dis_type: integer().
defining the similarity with optional values "FR_OSINE" or "FR_NORM_L2"

Return

retval: double

Python prototype (for reference only):

match(face_feature1, face_feature2[, dis_type]) -> retval

match(self, face_feature1, face_feature2, opts)

@spec match(
  t(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in(),
  [{:dis_type, term()}] | nil
) ::
  number() | {:error, String.t()}

Calculating the distance between two face features

Positional Arguments

self: Evision.FaceRecognizerSF.t()
face_feature1: Evision.Mat.
the first input feature
face_feature2: Evision.Mat.
the second input feature of the same size and the same type as face_feature1

Keyword Arguments

dis_type: integer().
defining the similarity with optional values "FR_OSINE" or "FR_NORM_L2"

Return

retval: double

Python prototype (for reference only):

match(face_feature1, face_feature2[, dis_type]) -> retval