https://bugs.kde.org/show_bug.cgi?id=464266
--- Comment #21 from caulier.gil...@gmail.com --- Hi, As all the face detection/recognition uses OpenCV in background (the neural network is based on this), the famous OpenCL option enable or not the support of GPU in OpenCV computation. If you look in OpenCV doc, all is explained. https://opencv.org/opencl/ There are others GPU option in OPenCV, for ex with NVIDIA cards. About the Apple, the GPU is proprietary, so, i'm not sure that OpenCV can already support this. In all cases, there is NO plan to introduce Pythin code in digiKam. This is a C++ project, and it's already difficult to maintain 1.5M of code lines, so to support Python, no way. The DNN used in face engine is based on OpenCV. So the solution is located in this library (the code must be compiled with GPU support in your computer, as all is optional in OpenCV). If there is other GPU options from OpenCV to support, sure this must be done. You are invited to contribute and to create a merge resquest over the digiKam code as explained in contribute page : https://www.digikam.org/contribute/ Best regards Gilles Caulier -- You are receiving this mail because: You are watching all bug changes.
