Dlib FaceLandmark Detector 1.2.8 (Unity Asset) [WiN x64]

By | August 12, 2022

 

Year/Date of Issue:22.05.2019
Version: 1.2.8
Developer: Enox Software
Developer site: _https://assetstore.unity.com/packages/tools/integration/dlib-facelandmark-detector-64314
Bit depth: 64bit
Interface language: English
Tabletka: not required
System requirements: Unity 5.6.6 or higher


Description:
Works with Unity Cloud Build
iOS and Android support
Windows 10 UWP support
WebGL support
Win & Mac and Linux standalone support
Editor preview support
DlibFaceLandmark Detector can use ObjectDetection and ShapePrediction using the Dlib19.7 C++ library.

Peculiarities:
– You can detect frontal human faces and face landmark (68 points, 17 points, 6 points) in Texture2D, WebCamTexture and Image array. In addition, you can detect other objects by modifying the trained data file.
– ObjectDetector is made using the now classic Histogram Oriented Gradients (HOG) function combined with a linear classifier, an image pyramid and a sliding window detection scheme. You can train your own detector in addition to the human face detector. If you would like to train your own detector, please refer to this page.
– ShapePredictor is created using the dlib implementation (“One millisecond face alignment with an ensemble of regression trees” by Waheed Kazemi and Josephine Sullivan, CVPR 2014). You can train your own models in addition to the human face landmark model using the dlib machine learning tools. If you would like to train your own models, please refer to this page.
– Extended examples using “OpenCV for Unity” are included. (These examples require “OpenCV for Unity”.)


Leave a Reply