vs2015
opencv>=3.0
CUDA (adaptive to your personal computer's enviroment)
cuDNN (version related to CUDA)
[1] 使用 CuDNN 进行卷积运算
[2] CUDNN 7.6.3 Docs
only use cuDNN&CUDA to build the VGG16 forward model for test, while model's parameters need be loaded.
C++
vs2015
opencv>=3.0
CUDA (adaptive to your personal computer's enviroment)
cuDNN (version related to CUDA)
[1] 使用 CuDNN 进行卷积运算
[2] CUDNN 7.6.3 Docs