Visage Technologies strives to minimize changes in API and configuration files when releasing new versions of the SDK. The inevitable changes are listed here, together with specific instructions for developers who have the existing applications built with older versions.

For each visage|SDK release, the incremental changes in relation to the previous release are listed. To apply the changes correctly, apply them in order from the older version to the newer version without skipping intermediary versions.

Contents


visage|SDK 8.8b2

General

The default legacy algorithm that could be allowed via the use_vnn configuration parameter (value 0) has been removed. The VNN algorithm is now the default and only tracking and detection algorithm.

The algorithm can be configured to improve feature points precision and robustness, over tracking speed (performance) by setting the refine_landmarks parameter to 1 in the configuration file. Otherwise, when performance is preferred over feature points precision, the recommendation is to disable refine_landmarks parameter by setting it to 0 in the configuration file.


Face tracking and detection algorithms are enhanced so that they can track and detect faces wearing protective masks of various colors and patterns.


Landmark detection changes:

Visible contour landmarks from group 13 (13.1-13.17) are no longer available (not detected or estimated). Instead, contour landmarks are now provided only in the form of a physical contour within group 15 (15.1-15.17). For more details please see FDP documentation.


Changing the detection of contour points from visible to physical results in improved stability and accuracy of the 3D head-pose estimation.


Changes in configuration file


ParametersFiles

use_vnn

Facial Features Tracker - Ultra.cfg

lbf_stage_modifier


lbf_nperturb


lbf_nperturb_threads



For detailed description of these changes, consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters.



  • refine_landmarks parameter

Parameter value

Effect

0

tracking algorithm in fast mode, less precise

1

tracking algorithm with the highest feature points precision, accuracy and robustness


For detailed description of these changes, consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters.




Configuration fileEffect

Facial Features Tracker - High.cfg

refine_landmarks parameter set to 1 (enabled)

Facial Features Tracker - Low.cfg

refine_landmarks parameter set to 0 (disabled)

Head Tracker.cfg 

refine_landmarks parameter set to 0 (disabled)


For detailed description of these changes, consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters.




Data files changes

As a consequence of improving algorithms, removing legacy algorithm, there are certain changes in the data files.

API

Status

Location

Files/Folders

VisageTracker
VisageFeaturesDetector

bdtsdata/NN/

fa.lbf, fc.lbf, is.bin, model.vino.bin, model.vino.xml

VisageTracker
VisageFeaturesDetector

bdtsdata/FF/

ff.dat

VisageTracker
VisageFeaturesDetector

bdtsdata/LBF/

lv

VisageTracker
VisageFeaturesDetector

bdtsdata/NN/vnn

std_dev_image.bin, mean_image.bin

VisageTracker
VisageFeaturesDetector

bdtsdata/NN/vnnbdtsdata/NN/vnn


Projects using older versions of these files should be updated to contain the newest data files from the bdtsdata folder.


Sample changes

Samples displaying feature points are updated to draw physical contour points.

visage|SDK 8.7

General

The in-house developed runner is no longer available and is being replaced by OpenVINO™ toolkit, which is now the only and default neural network runner.

OpenVINO™ toolkit significantly improves the performance of visage|SDK algorithms.

It is implemented in the following libraries:

  • OVPlugin.dll,

  • inference_engine.dll,

  • MKLDNNPlugin.dll,

  • mkl_tiny_tbb.dll,

  • tbb.dll

  • libmmd.dll,

  • svml_dispmd.dll

These libraries are dependencies of libVisageVision64.dll and should now be included in projects along with other visage|SDK libraries. Additionally, OpenVINO™ toolkit requires its own set of data files with extensions .vino.bin and .vino.xml provided in Samples/data/bdtsdata.

OpenVINO is a trademark of Intel Corporation or its subsidiaries.


The old face recognition model is being replaced by a smaller, faster and more accurate face recognition model.


Introducing a more accurate and robust face detection model. The new model is used in face tracking and face detection when use_vnn configuration parameter is set to 1. Otherwise, the previous face detection model will be used.

For more information about using VNN detection algorithm, please consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters and VisageFeaturesDetector documentation.



VNN algorithm can now be configured for better performance, at the cost of feature points precision. This mode is enabled by setting use_vnn parameter in configuration file to 1.

Recommended for usage when performance is preferred over feature points precision, e.g. when interested in fast performing head pose (head rotation and translation) estimation.


API changes

Added support for multiple image format for extractDescriptor() and addDescriptor() methods.

extractDescriptor()

In addition to RGB and grayscale, now accepts RGBA input image

addDescriptor()

In addition to RGB, now accepts RGBA and grayscale input image


Changes in configuration file

use_vnn configuration parameter added to VisageFeaturesDetector configuration file FaceDetector.cfg(located in Samples/data/bdstdata) and set to value 1.

VNN algorithm will be used in VisageFeaturesDetector API by default.

use_vnn configuration parameter values changed and additional value added:

Parameter value

Effect

0

VNN algorithm is disabled, default tracking algorithm and detection model are used

1

VNN algorithm enabled in fast mode, at the cost of feature points precision

2

VNN algorithm enabled, improved precision, accuracy and robustness


For detailed description of these changes, consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters.

If you want to update your existing configuration files, it is recommended to copy the parameters values from Facial Features Tracker - Ultra.cfg configuration file supplied in this package.


Data files changes

As a consequence of improving algorithms, removing in-house developed runner, and improving VNN algorithm, there are certain changes in the data files.

API

Status

Location

Files/Folders

VisageFaceRecognition

bdtsdata/NN/

fr.bin

VisageFaceAnalyser

bdtsdata/LBF/vfadata/ad/

ae.bin

VisageTracker
VisageFeaturesDetector

bdtsdata/NN/

pr.bin, model.bin

VisageFaceRecognition

bdtsdata/FF/vnn/

fr.vino.bin, fr.vino.xml

VisageTracker
VisageFeaturesDetector

bdtsdata/FF/vnn/

ff.vino.bin, ff.vino.xml

VisageTracker
VisageFeaturesDetector

bdtsdata/NN/vnn

bdtsdata/NN/vnn


Projects using older versions of these files should be updated to contain the newest data files from the bdtsdata folder.


Sample changes

VisageTrackerUnityPlugin

Function _grabFrame() captures RGB image, instead of BGR image.

VisageTrackerUnityDemo

The previously used BGRATex texture shader for applying captured frame image data was replaced with RGBATex texture shader (Tracker.cs script).


visage|SDK 8.6.1

General

To improve the performance of our algorithms and to support a wider variety of neural network models, we are introducing a configurable framework for choosing between different neural network runners.

As a result, additional configuration file NeuralNet.cfg is now included in visage|SDK (located in Samples/data). This file allows the users to configure which runner will be used by visage|SDK. Users can choose between:

  • Visage Technologies' runner developed in-house and

  • OpenVINO™ toolkit.

For more information on the parameters of NeuralNet.cfg file see the API page.

OpenVINO is a trademark of Intel Corporation or its subsidiaries.


New experimental algorithm for face tracking and alignment introduced - VNN algorithm.

For the price of slightly reduced tracking speed/performance, it significantly improves tracking quality and robustness. VisageTracker and VisageFeaturesDetector can be configured to use VNN algorithm via configuration parameter - use_vnn.

For more information on use_vnn parameter, please consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters and VisageTracker class documentation.


It is recommended to use VNN algorithm with OpenVINO™ toolkit which significantly improves the speed of running neural networks, thus mitigating any performance reductions.

OpenVINO is a trademark of Intel Corporation or its subsidiaries.


API changes

Introducing new C++ and C# API for programmatically changing VisageTracker configuration parameters, including new additional classes and templates:

VisageConfiguration

Modify configuration parameters on the fly

VsCfgArr

Helper template structure for storing various VisageConfiguration array data types

The aforementioned classes are used in conjunction with new VisageTracker methods:

C++

VisageConfiguration VisageTracker::getTrackerConfiguration()

void VisageTracker::setTrackerConfiguration(VisageConfiguration &configuration)

C#

VisageConfiguration VisageCSWrapper.VisageTracker.GetTrackerConfiguration()

void VisageCSWrapper.VisageTracker.SetTrackerConfiguration(VisageConfiguration configuration)


There are slight differences in usage between C++ and C#. For example, C++ API uses helper structure VsCfgArr to return specific data types where C# uses native C# types.


VisageConfiguration and VsCfgArr class documentation contains more details and examples of usage in code.



FDP group 10 (ears) has been extended from 10 to 24 points (12 points per ear) as part of the ear tracking feature.

FDP files saved with visage|SDK 8.6 will not be backwards compatible with the previous versions due to the addition of new FDP points.

Using FDP::readFromFile() to load an 8.6 FDP file in an earlier version of the visage|SDK will lead to undefined behavior.



VisageTracker method stop has been deprecated from both APIs.

C++

VisageTracker::stop()

C#

VisageCSWrapper.VisageTracker.Stop()



Prototype of method

void initializeLicenseManager(const char *licenseKeyFileName)

changed to

int initializeLicenseManager(const char *licenseKeyFileName)

It is no longer necessary to declare the licensing function prototype explicitly within your code. Including any of the following headers will also include the licensing prototype:

  • VisageTracker.h

  • VisageFeaturesDetector.h

  • VisageFaceRecognition.h

  • VisageFaceAnalyser.h



FeaturePoint class and FDP class have additional property and functions, respectively to conform with the C++ API.

FeaturePoint

FeaturePoint.detected

FDP

bool FDP::FPIsDetected(int group, int n)
bool FDP::FPIsDetected(String name)


1 indicates that the feature point is obtained from a 2D image using the tracking algorithm. 0 indicates that the feature point is estimated from fitting a 3D model onto the detected feature points of the face.


Changes in configuration file

Two additional configuration files have been added. One for the ear tracking feature and one for the novel tracking algorithm.

Configuration name

Parameter changed/added

Parameter value

Facial Features Tracker - High - With Ears.cfg

efine_ears
*_fitting_model
*_fitting_fdp

1
jk_300_wEars.wfm
jk_300_wEars.fdp

Facial Features Tracker - Ultra.cfg

use_vnn

1


For detailed description of these changes consult VisageTracker Configuration Manual, paragraph 1.1.Standard configuration files

If you want to update your existing configuration files, it is recommended to copy the parameters values from Facial Features Tracker - Ultra.cfg configuration file supplied in this package.



  • refine_ears parameter added, off (0) by default. Toggles the tracking and refinement of ear points (group 10) for VisageTracker and VisageFeaturesDetector.

Used together with ears 3D model - jk_300_wEars.wfm

  • use_vnn parameter added, off (0) by default. Toggles usage of the new experimental algorithm for face tracking and alignment - VNN algorithm

For detailed description of these changes, consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters.


If you want to update your existing configuration files, it is recommended to copy the parameters values from Facial Features Tracker - High.cfg configuration file supplied in this package.


3D Model changes

A new model file has been added for ear tracking functionality - jk_300_wEars.wfm. The model contains an additional 334 polygons and its vertices are mapped to 14 new FDP points in group 10 (10.11 - 10.24).

For detailed description of these changes, consult VisageTracker Configuration Manual, paragraph 2.3.The 3D models used in tracking


Data files changes

New data files and model files for ear tracking added

Location

Files

Samples/data/bdtsdata/NN

efa.lbf
efc.lbf

Samples/data/

jk_300_wEars.wfm
jk_300_wEars.fdp



New data folder and data files added for VNN algorithm in Samples/data/bdtsdata/NN/vnn

Projects using older versions of these files should be updated to contain the newest data files from the bdtsdata folder.


Sample changes

FacialAnimationUnityDemo sample application has been removed and will no longer be distributed.


VisageTrackerUnityDemo sample application is distributed as a Unity project, not as a prebuilt application.

Build and run instructions are provided in the VisageTrackerUnityDemo sample documentation.



Instructions on visage|SDK integration with Unity can be found here.






visage|SDK 8.5

Changes in configuration file

  • smoothing_factors parameter
    Due to the re-implementation of the smoothing algorithm in VisageTracker, default values and optimal ranges for this parameter have been changed in all configurations.

Please consult VisageTracker Configuration Manual, paragraph 2.1.Configuration parameters for additional information.


If you want to update your existing configuration files, it is recommended to copy the parameters values from Facial Features Tracker - High.cfg configuration file supplied in this package.


Sample changes

VisageRendering.cs:

Methods DisplayEmotion() and DisplayAgeGenderName() have changed the prototypes from:

public void DisplayEmotion(FaceData trackingData, float[] emotions, int width, int height,
                           bool face_recognition_mode=false)
public void DisplayAgeGenderName(FaceData trackingData, float age, int gender, string name, 
                                 int width, int height, int drawingOptions, 
                                 bool face_recognition_mode=false)

to:

public void DisplayEmotion(FaceData trackingData, float[] emotions, int width, int height, 
                           bool display_background = false)
public void DisplayAgeGenderName(FaceData trackingData, float age, int gender, string name, 
                                 int width, int height, bool display_background = true)


Build tools changes

Libraries built with msvc100 are no longer supported within the package.

Data files changes

visage|SDK data files located in Samples/data/bdtsdata folder have been updated.

Copy all files located in the Samples/data/bdtsdata folder to appropriate folders in any existing application.