This is a branch of Caffe supporting R-FCN, which has been tested under Windows (Windows 7, 8, Server 2012 R2) and Linux (Ubuntu 14.04).
Copy Makefile.config.example to Makefile.config
We need to modify Makefile.config to specify some software PATHS, you may view my Makefile.config for reference.
Download CUDA Toolkit 7.5 from nVidia website.
Uncomment MATLAB_DIR and set MATLAB_DIR accordingly to build Caffe Matlab wrapper. Matlab 2014a and later versions are supported.
For cuDNN acceleration using NVIDIA’s proprietary cuDNN software, uncomment the USE_CUDNN := 1 switch in Makefile.config. cuDNN is sometimes but not always faster than Caffe’s GPU acceleration.
Download cuDNN v3 or cuDNN v4 from nVidia website. And unpack downloaded zip to $CUDA_PATH (It typically would be /usr/local/cuda/include and /usr/local/cuda/lib64).
Simply type
make -j8 && make matcaffe
Requirements: Visual Studio 2013
Copy .\windows\CommonSettings.props.example to .\windows\CommonSettings.props
3rd party dependencies required by Caffe are automatically resolved via NuGet.
Download CUDA Toolkit 7.5 from nVidia website.
Set MatlabSupport to true and MatlabDir to the root of your Matlab installation in .\windows\CommonSettings.props to build Caffe Matlab wrapper. Matlab 2014a and later versions are supported.
Download cuDNN v3 or cuDNN v4 from nVidia website.
Unpack downloaded zip to %CUDA_PATH% (environment variable set by CUDA installer).
Alternatively, you can unpack zip to any location and set CuDnnPath to point to this location in .\windows\CommonSettings.props.
CuDnnPath defined in .\windows\CommonSettings.props.
By default, cuDNN is not enabled. You can enable cuDNN by setting UseCuDNN to true in the property file.
Now, you should be able to build .\windows\Caffe.sln
After you have built solution with Matlab support, copy all files in .\Build\x64\Release to R-FCN\external\caffe\matlab\caffe_rfcn.
Refer to the BVLC/caffe master branch README for all other details such as license, citation, and so on.