General

Before you can call the SDK functions, you must first initialize the SDK with your desired configuration options. These configuration options will ultimately impact the behaviour of the SDK, so be sure to read through what each one does carefully.

Once you have initialized the SDK, then proceed to the the License section.

Important

By default, the SDK is initialized with the tfsdk.FACIALRECOGNITIONMODEL.LITE face recognition model. This is our fastest and most lightweight face recognition model; however, it is not the most accurate face recognition model we offer. To use a more accurate model, refer to the tfsdk.FACIALRECOGNITIONMODEL section below then be sure to initialize the SDK with the desired tfsdk.ConfigurationOptions.

SDK.__init__(*args, **kwargs)

Overloaded function.

  1. __init__(self: tfsdk.SDK) -> None

Initialize the SDK using default configuration options.

  1. __init__(self: tfsdk.SDK, configuration_options: tfsdk.ConfigurationOptions) -> None

    Initialize the SDK using custom configuration options.

    Parameters

    configuration_options - custom configuration options.

static SDK.get_version() → str

Gets the version-build number of the SDK.

Returns

The SDK version.

class tfsdk.FACIALRECOGNITIONMODEL

Facial recognition models. To compare model performances, refer to our ROC curves. You can also find more information on our FAQ page. The current most accurate model is TFV5.

Members:

LITE : Lightweight model with faster inference time but reduced accuracy, ideal for embedded systems or lightweight CPU only deployments, and for 1 to 1 matching use cases.

FULL : Full TFV4 model which has better accuracy than the LITE model, but also has greater inference time. Ideal for GPU deployments and for 1 to N use cases. Note, TFV4 has now been deprecated and replaced by TFV5 which has better performance. Despite this, we will continue providing support for TFV4 for clients with existing collections.

TFV5 : TFV5 is currently the highest accuracy model. Ideal for GPU deployments and for 1 to N use cases.

class tfsdk.OBJECTDETECTIONMODEL

Object detection models

Members:

ACCURATE

FAST

class tfsdk.FACEDETECTIONMODE

Face detection modes

Members:

ROBUST : The ROBUST mode is optimized for face recognition, smallestFaceHeight less than 80 pixels will be overridden.

VERSATILE : The VERSATILE mode can detect small faces (as small as 16x16 pixel) and can be faster than the ROBUST mode.

class tfsdk.FACEDETECTIONFILTER

Face detection filter

Members:

HIGH_RECALL : The detected face scores are thresholded to have a low precision and a high recall rate.

HIGH_PRECISION : The detected face scores are thresholded to have a high precision and a low recall rate.

BALANCED : The detected face scores are thresholded to have a medium precision and recall rate.

UNFILTERED : The detected face scores are not thresholded.

class tfsdk.DATABASEMANAGEMENTSYSTEM

Database Management System for storing Faceprints

Members:

SQLITE : Use sqlite backend. Write Faceprints to local disk. Ideal for embedded systems or use cases where only one process connects to the database.

POSTGRESQL : Use PostgreSQL backend. Ideal for distributed systems requiring synchronization.

NONE : Do not write Faceprints to disk, only store in ram. Warning, enrolled Faceprints will not be saved after the program terminates. Switching to a new collections will also delete all enrolled templates.

class tfsdk.EnableGPU

Enable GPU support (default is False). Note, GPU support requires a different version of the SDK. You are able to enable GPU for all modules or only specific modules.

property face_detector

Enable GPU inference for face detection

property face_recognizer

Enable GPU inference for face recognition template generation

EnableGPU.__init__(*args, **kwargs)

Overloaded function.

  1. __init__(self: tfsdk.EnableGPU) -> None

Default constructor initializes GPU inference for all modules to false

  1. __init__(self: tfsdk.EnableGPU, val: bool) -> None

    Enable or disable GPU inference for all modules.

    Parameters

    val - enable GPU for all modules.

class tfsdk.ConfigurationOptions
property GPU_device_index

GPU device index (default is 0)

property dbms

Database management system for storing Faceprints (default is SQLITE). See DATABASEMANAGEMENTSYSTEM

property enable_GPU

Enable GPU support (default is False). Note, GPU support requires a different version of the SDK. You are able to enable GPU for all modules or only specific modules. See EnableGPU

property fd_filter

The face detection precision-recall mode (default is BALANCED). See FACEDETECTIONFILTER

property fd_mode

The face detection mode (default is VERSATILE). See FACEDETECTIONMODE

property fr_model

The model to be used for facial recognition (default is LITE model). See FACIALRECOGNITIONMODEL

property fr_vector_compression

Improves the computation speed for 1 to 1 comparisons and 1 to N searches. Also reduces the feature vector length (default is False).

property models_path

The directory path containing the model files (default is ./ )

property obj_model

The model to be used for object detection (default is ACCURATE model). See OBJECTDETECTIONMODEL

property smallest_face_height

The smallest face height that the face detector can detect (default is 40 pixels). The face detector has a detection scale range of about 5 octaves. Ex. 40 pixels yields the detection scale range of ~40 pixels to 1280 (=40x2^5) pixels. If set to -1, will dynamically adjusts the face detection scale range from image-height/32 to image-height to ensure that large faces are detected in high resolution images.

class tfsdk.ERRORCODE

Members:

NO_ERROR

INVALID_LICENSE

FILE_READ_FAIL

UNSUPPORTED_IMAGE_FORMAT

UNSUPPORTED_MODEL

NO_FACE_IN_FRAME

COLLECTION_CREATION_ERROR

DATABASE_CONNECTION_ERROR

ENROLLMENT_ERROR

FAILED

NO_RECORD_FOUND

NO_COLLECTION_FOUND

MAX_COLLECTION_SIZE_EXCEEDED

COLLECTION_DELETION_ERROR