如何在具有CUDA的OpenCV中加快new.forward（）进程？

Question

在具有CUDA的Python中，如何在OpenCV中加快new.forward()进程？

我写

net.setPreferableTarget(cv2.dnn.DNN_TARGET_OPENCL)

我的速度慢了一倍 并显示下一条消息作为输出：

OpenCV(ocl4dnn): consider to specify kernel configuration cache directory 
                 via OPENCV_OCL4DNN_CONFIG_PATH parameter.
OpenCL program build log: dnn/dummy
Status -11: CL_BUILD_PROGRAM_FAILURE
-cl-no-subgroup-ifp
Error in processing command line: Don't understand command line argument "-cl-no-subgroup-ifp"!

我有GeForce GTX 1050和Cuda 10