First time here? Check out the FAQ!
cuda::DescriptorMatcher::knnMatch runs in differ I want to use cuda::DescriptorMatcher::knnMatch to speed up my feature