I'm interested in incorporating TensorFlow into a C++ server application built in Visual Studio on Windows 10 and I need to know if that's possible.
Google recently announced Windows support for TensorFlow: https://developers.googleblog.com/2016/11/tensorflow-0-12-adds-support-for-windows.html
but from what I can tell this is just a pip install for the more commonly used Python package, and to use the C++ API you need to build the repo from source yourself: How to build and use Google TensorFlow C++ api
I tried building the project myself using bazel, but ran into issues trying to configure the build.
Is there a way to get TensorFlow C++ to work in native Windows (not using Docker or the new Windows 10 Linux subsystem, as I've seen others post about)?
Thanks,
Ian
It is certainly possible to use TensorFlow's C++ API on Windows, but it is not currently very easy. Right now, the easiest way to build against the C++ API on Windows would be to build with CMake, and adapt the CMake rules for the tf_tutorials_example_trainer project (see the source code here). Building with CMake will give you a Visual Studio project in which you can implement your C++ TensorFlow program.
Note that the tf_tutorials_example_trainer project builds a Console Application that statically links all of the TensorFlow runtime into your program. At present we have not written the necessary rules to create a reusable TensorFlow DLL, although this would be technially possible: for example, the Python extension is a DLL that includes the runtime, but does not export the necessary symbols to use TensorFlow's C or C++ APIs directly.
There is a detailed guide by Joe Antognini and a similar TensorFlow ReadMe at GitHub explaining the building of TensorFlow source via CMake. You also need to have SWIG installed on your machine which allows connecting C/C++ source with the Python scripting language. I did use Visual CMAKE (cmake-gui) with the screen capture shown below.
In the CMake configuration, I used Visual Studio 15 2017 compiler. Once this stage successfully completes, you can click on the Generate button to go ahead with the actual build process.
However, on Visual Studio 2015, when I attempted building via the "ALL_BUILD" project, the setup gave me "build tools for v141 cannot be found" error. This did not go away even when I attempted to retarget my solution. Finally, the solution got built successfully with Visual Studio 2017. You also need to manually set the SWIG_EXECUTABLE path in CMake before it successfully configures.
As indicated in the Antognini link, for me the build took about half an hour on a 16GB RAM, Core i7 machine. Once done, you might want to validate your build by attempting to run the tf_tutorials_example_trainer.exe file.
Hope this helps!
For our latest work on building TensorFlow C++ API on Windows, please look at this github page. This works on Windows 10, currently without CUDA support (only CPU).
PS:
Only the bazel build method works, because CMake is not supported and not maintained anymore, resulting in CMake configuration errors.
I had to use a downgraded version of my Visual Studio 2017 (from 15.7.5 to 15.4) by adding "VC++ 2017 version 15.4 v14.11 toolset" through the installer (Individual Components tab).
The cmake command which worked for me was:
cmake .. -A x64 -DCMAKE_BUILD_TYPE=Release ^
-T "v141,version=14.11" ^
-DSWIG_EXECUTABLE="C:/Program Files/swigwin-3.0.12/swig.exe" ^
-DPYTHON_EXECUTABLE="C:/Program Files/Python/python.exe" ^
-DPYTHON_LIBRARIES="C:/Program Files/Python/libs/python27.lib" ^
-Dtensorflow_ENABLE_GPU=ON ^
-DCUDNN_HOME="C:/Program Files/cudnn-9.2-windows10-x64-v7.1/cuda" ^
-DCUDA_TOOLKIT_ROOT_DIR="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v9.0"
After the build, open tensorflow.sln in Visual Studio and build ALL_BUILD.
If you want to enable GPU computation, do check your Graphics Card here (Compute Capability > 3.5). Do remember to install all the packages (Cuda Toolkit 9.0, cuDNN, Python 3.7, SWIG, Git, CMake...) and add the paths to the environment variable in the beginning.
I made a README detailing how to I built the Tensorflow dll and .lib file for the C++ API on Windows with GPU support building from source with Bazel. Tensorflow version 1.14
The tutorial is step by step and starts at the very beginning, so you may have to scroll down past steps you have already done, like checking your hardware, installing Bazel etc.
Here is the url: https://github.com/sitting-duck/stuff/tree/master/ai/tensorflow/build_tensorflow_1.14_source_for_Windows
Probably you will want to scroll all the way down to this part:
https://github.com/sitting-duck/stuff/tree/master/ai/tensorflow/build_tensorflow_1.14_source_for_Windows#step-7-build-the-dll
It shows how to pass command to create .lib and .dll.
Then to test your .lib you should link it into your c++ project,
Then it will show you how to identify and fix the missing symbols using the TF_EXPORT macro
I am actively working on making this tutorial better so feel free to leave comments on this answer if you are having problems.
I recently use the pre-built OpenCV 3.1.0 on Windows which was downloaded from here. Actually, I followed the official installation.
The thing is that I find that the VideoCapture module of pre-built OpenCV processes video very slowly. It seems that it has no support of ffmpeg. And I find the official note:
To use the OpenCV library you have two options: Installation by Using the Pre-built Libraries or Installation by Making Your Own Libraries from the Source Files . While the first one is easier to complete, it only works if you are coding with the latest Microsoft Visual Studio IDE and doesn't take advantage of the most advanced technologies we integrate into our library.
It makes me curious about what is the actual build configuration of pre-built OpenCV 3.1.0 (or other versions) on Windows. No supports of TBB, IPP, Eigen, CUDA, etc...? I didn't find any clue on the internet. Anyone knows?
I'm trying to write a program that uses SURF algorithm and I know that the nonfree module must be installed separately. I've downloaded and installed the latest version of CMake(3.5.2) and I'm following the instructions from:
https://github.com/itseez/opencv_contrib/ . I'm using the GUI and I run Visual Studio 2015 on a 64 bit Windows 10.
Since I know that for SURF you must include xfeatures2d when it asked me the source of the code I only specified the xfeatures2d folder. When I first pressed the configure button I had some errors(I've attached an image of them). I managed to get rid of one of the errors, the one that said to write a line of code at the top of the file
cmake_minimum_required(VERSION 3.5), but I still have one error:
CMake Error at CMakeLists.txt:4 (ocv_define_module):
Unknown CMake command "ocv_define_module".
I'm also attaching a picture of the CMakeLists.txt. Please help me find the problem.
Also, if you could help my install all modules at once, I would be grateful. Or do I have to set as input every folder in the modules folder?
GUI error and CMakeLists.txt
Try using Visual Studio 12 2013 for compilation, this has worked for me, but not the other (newer) versions of Visual Studio.
I want to cross compile opencv with ffmpeg for arm.
I have cross compiled ffmpeg, how do I configure cmake to see the ffmpeg?
Edit: I am cross compiling for AR Drone 2.0. I use toolchain under platform/linux.
The answer to your specific problem is to use cmake-gui which provides you a visual interface to ease the process of setting up variables, like your ffmpeg package location.
However I have 2 recommendations:
Download the most updated toolchain.
DO NOT USE THIS GUIDE. Use instead the toolchain that you find in the package in sources\platforms\android\android.toolchain.cmake: read inside, you'll find the correct instructions at the beginning of the file.
Another optional recommendation: use ninja as build system, much faster.
I was successful in cross compiling OpenCV with ffmpeg for AR Drone's ARM processor by writing my own toolchain file where I explicitly set all the linker, pkg_config and compiler paths to relevant directories containing the cross-compiled stuff.
I have written a blog post detailing the entire process here:
How to Cross-Compile OpenCV with FFmpeg for AR Drone (ARM Processor).
Does anybody know how to build cvBlobsLib using MinGW? On official page http://opencv.willowgarage.com/wiki/cvBlobsLib there is only instruction for VS.
There is also linux version of this lib http://opencv.willowgarage.com/wiki/cvBlobsLib?action=AttachFile&do=view&target=cvblobs8.3_linux.tgz , but its makefile cannot be used in windows as i see.
If you use eclipse then you dont have a lot of work:
Create a new project, using MinGW toolchain.
Go to the project properties, and under C/C++ General >> Paths and Symbols add the openCV library paths.
compile the project and it should be OK.
Use this
http://opencv.willowgarage.com/wiki/cvBlobsLib#Build_intructions
if you have more problems (especially NOTE 3)