DlibFaceLandmarkDetector is an Assets Plugin for object detection and shape prediction using the Dlib C++ library within the Unity.DlibFaceLandmarkDetector is an Assets Plugin for object detection and shape prediction using Dlib19.7 C++ Library.Official Site | ExampleCode | Android Demo WebGL Demo | Tutorial & Demo Video | Forum | API ReferenceOUR ASSET FEATURESCross PlatformiOS & Android & Windows10 UWP support.Win & Mac & Linux Standalone support.WebGL support.ChromeOS support.visionOS support.(beta)Support for preview in the Editor.Object Detection & Shape Prediction- Object Detection: Detect frontal human faces and other objects using Histogram of Oriented Gradients (HOG) feature combined with a linear classifier, an image pyramid, and sliding window detection scheme.- Shape Prediction: Detect face landmarks (68 points, 17 points, 6 points) using dlib's implementation of the paper (One Millisecond Face Alignment with an Ensemble of Regression Trees by Vahid Kazemi and Josephine Sullivan, CVPR 2014).- Custom Training: You can train your own object detector and shape predictor models using dlib's machine learning tools. (Object Detector Training Tool, Shape Predictor Training Tool)Easy to Use- Support for Texture2D , WebCamTexture , and Image byte array input.- When combined with OpenCV for Unity, you can also input from OpenCV's Mat class for enhanced image processing capabilities.- Helper functions for easy integration with Unity.- FaceLandmarkDetector class implement Mat class for IDisposable , allowing you to manage resources using the using statement.Include Many Examples- Includes a wide variety of example usage scenarios, which consist of scene files and script codes. By running these example applications, you can learn how to develop Dlib applications effectively.- Basic Examples: Texture2D Example, WebCamTexture Example, Benchmark Example.- Advanced Examples: Advanced examples using OpenCV for Unity. (requires OpenCV for Unity)DlibFaceLandmarkDetector Examples (GitHub)EnoxSoftware repositories (GitHub)Visual Scripting SupportYou can take full advantage of all the features of DlibFaceLandmarkDetector in Unity’s Visual Scripting. You can learn how to integrate them through the VisualScripting With DlibFaceLandmarkDetector Example (GitHub).Demo:- Android Demo- WebGL DemoExamples:Texture 2DExampleWebCamTexture ExampleBenchmark ExampleAdvanced Examples:(require OpenCV for Unity)Texture2DToMat ExampleWebCamTexture2MatHelper ExampleVideoCapture2MatHelper ExampleARHead WebCamTexture ExampleARHead VideoCapture ExampleFrameOptimization ExampleNoiseFilter WebCamTexture ExampleNoiseFilter VideoCapture ExampleExampleCode using Dlib FaceLandmark Detector is available.FaceMask ExampleHoloLens With DlibFaceLandmarkDetector ExampleDlibFaceLandmarkDetector uses Dlib under Boost Software License ; see Third-Party Notices.txt file in package for details.The Shape Predictor model files included with this asset are available for commercial use.System RequirementsBuild Win Standalone & Preview Editor : Windows 8 or laterBuild Mac Standalone & Preview Editor : OSX 10.13 or laterBuild Linux Standalone & Preview Editor : Ubuntu18.04 or laterBuild Android : API level 21 or laterBuild iOS : iOS Version 12.0 or laterBuild VisionOS : visionOS 1 or later (beta)More Info >>Supported OS iOS, Android, ChromeOS, Windows10 UWP, WebGL, Windows, macOS, Linux, visionOS (beta)Link to documentation API Reference
DlibFaceLandmarkDetector は、Dlib C++ ライブラリを使用してオブジェクト検出と形状予測を行うことができます。
詳細
Works with Unity Cloud Build
ChromeOS support
iOS & Android support
Windows10 UWP support
Lumin ( MagicLeap ) support support
WebGL support
Win & Mac & Linux Standalone support
Preview support in the Editor
DlibFaceLandmarkDetectorはDlib19.7 C++ Libraryを使用してObjectDetectionとShapePredictionをUnityで行うためのAssetプラグインです。
Official Site | ExampleCode | Android Demo WebGL Demo | Tutorial & Demo Video | Forum | API Reference
機能紹介:
- Texture2D、WebCamTexture、画像バイト列からfrontal human facesとface landmark (68 points, 17points, 6points)を検出可能です。さらに、別の訓練データファイルに変更することによって、任意のオブジェクトを検出することが出来ます。
- ObjectDetectorはHOG特徴量を用いて物体を検出します。 同梱の顔の検出器だけではなく、自分で訓練した物体検出器を使用することも可能です。詳細はこちらのページを参照してください。
- ShapePredictorはこの論文を元に実装されています(One Millisecond Face Alignment with an Ensemble of Regression Trees by Vahid Kazemi and Josephine Sullivan, CVPR 2014)。 同梱の顔のランドマークの形状推定モデルだけではなく、dlibの学習ツールを使用して自分で形状推定モデルを訓練して使用することが可能です。詳細はこちらのページを参照してください。
- “OpenCV for Unity”を利用した高度なサンプルも同梱されています。（“OpenCV for Unity”が必要です。）
- PlayMakerActions for DlibFaceLandmarkDetectorが利用可能です。
サンプル:
Texture2DExample
WebCamTextureExample
CatDetectionExample
OpenCVforUntiy利用サンプル:
(require OpenCV for Unity)
Texture2DToMatExample
WebCamTextureToMatHelperExample
VideoCaptureExample
ARHeadWebCamTextureExample
VideoCaptureARHeadExample
FrameOptimizationExample
NoiseFilterWebCamTextureExample
NoiseFilterVideoCaptureExample
Dlib FaceLandmark Detectorを使ったサンプルコードを公開中です。
FaceSwapper Example
FaceMask Example
HoloLens with DlibFaceLandmarkDetector Example
MagicLeap with DlibFaceLandmarkDetector Example
DlibFaceLandmarkDetector uses Dlib under Boost Software License ; see Third-Party Notices.txt file in package for details.
The Shape Predictor model files included with this asset are available for commercial use.
システム条件:
Build Win Standalone & Preview Editor : Windows 8 or later
Build Mac Standalone & Preview Editor : OSX 10.9 or later
Build Linux Standalone & Preview Editor : Ubuntu16.04 or later
More Info >>
