Tensorflow Faster R-CNN for Windows and Linux by using Python 3
This is the branch to compile Faster R-CNN on Windows and Linux. It is heavily inspired by the great work done here and here. I have not implemented anything new but I fixed the implementations for Windows, Linux and Python 3.
Currently, this repository supports Python 3.5, 3.6 and 3.7. Thanks to @morpheusthewhite
PLEASE
BE AWARE: I do not have time or intention to fix all the issues for
this branch as I do not use it commercially. I created this branch just
for fun. If you want to make any commitment, it is more than welcome.
Tensorflow has already released an object detection api. Please refer to
it. https://github.com/tensorflow/models/tree/master/research/object_detection
If you find a solution to an existing issue in the code, please send a PR for it.
Install tensorflow, preferably GPU version. Follow instructions.
If you do not install GPU version, you need to comment out all the GPU
calls inside code and replace them with relavent CPU ones.
Checkout this branch
Install python packages (cython, python-opencv, easydict) by running pip install -r requirements.txt (if you are using an environment manager system such as conda you should follow its instruction)
Go to ./data/coco/PythonAPI Run python setup.py build_ext --inplace Run python setup.py build_ext install Go to ./lib/utils and run python setup.py build_ext --inplace
Follow these instructions
to download PyCoco database.
I will be glad if you can contribute with a batch script to
automatically download and fetch. The final structure has to look like dataVOCDevkit2007VOC2007
Download pre-trained VGG16 from here and place it as dataimagenet_weightsvgg16.ckpt. For rest of the models, please check here