Version 1.3.8（current） [Common]Changed to use unsafe code by default. [Common]Optimized the amount of memory allocation, in the FaceLandmarkDetector class.

A new target platform has been added to our Unity Asset “OpenCVForUnity” and “DlibFaceLandmarkDetector“. With this update, it is now possible to develop WebGL applications that users can easily play with web browsers. Although it is currently in beta version, it seems that the main functions are working without problems. It works with Unity 5.3 or later. Demo Application Click To Start Development Tips Put a file that you want to use for the Utils.getFilePathAsync() in the “Aseets/StreamingAssets/” folder. In Case of WebGL platform, you need to use the Utils.getFilePathAsync() instead of the Utils.getFilePath(). (The haarcascade_frontalface_alt.xml is for OpenCVForUnityExample’s scenes. Please copy only when necessary) On the WebGL platform (asm.js), the calculation result of Float type may be significantly different from other platforms. When using the OpenCV’s method that use the Mat class (CvType is CV_32F) as an argument, you need to pay attention to the calculation precision. Please try experiencing WebGL development.

PlayMakerActions for DlibFaceLandmarkDetector 1.0.0 is now available.https://www.assetstore.unity3d.com/#!/content/95863