Convert caffe to tensorflow for Windows using Docker. Docker uses caffe-tensorflow (https://github.com/ethereon/caffe-tensorflow) as a kernel for convertation. Inspired by https://github.com/nicolov/segmentation_keras/tree/master/conversion.
Tested on Docker for Win10 (you may get Docker for Windows from here: https://docs.docker.com/docker-for-windows/install/#download-docker-for-windows). To run docker I used Windows PowerShell. Have not tested on previous versions of Windows yet. In this case you need a virtual machine for running docker (e.g. https://www.virtualbox.org/), and Docker Toolbox (https://docs.docker.com/toolbox/toolbox_install_windows/).
Add links to your caffe files you want to convert (.caffemodel
and .prototxt
).
Build a docker image from the Dockerfile:
docker build . -t newimage
Run the command:
docker run -v $pwd/converted:/workspace -ti newimage
The workspace of the container is stored in ./converted
.