ubuntu16.04+python3.5.0+cuda8.0+cudnn运行代码faster-rcnn-tensorflow-python3.5-master代码

１．代码：https://github.com/dBeker/Faster-RCNN-TensorFlow-Python3.5

２．先配置一个环境：

　　打开终端：conda create --name python350(你自己可以任意设置的环境名称)　python==3.5.0(在这个代码中一定要是3.5.0)

进入新环境：source activate python350

3.安装cython, python-opencv, easydict

4.到./data/coco/PythonAPI目录下打开终端：运行　python setup.py build_ext --inplace　　python setup.py build_ext install

5.下载数据集　https://mp.csdn.net/postehttps://github.com/rbgirshick/py-faster-rcnn#beyond-the-demo-installation-for-training-and-testing-models

6.下载vgg16 http://download.tensorflow.org/models/vgg_16_2016_08_28.tar.gz　,记得把vgg_16.ckpt改为vgg16.ckpt

7.运行train.py

注意：在运行train.py过程中提示缺少哪个模块，就install.

大概运行５０个小时

https://github.com/TuSimple/simpledet/tree/master/models/tridentnet

https://zhuanlan.zhihu.com/p/54334986?utm_source=wechat_session&utm_medium=social&utm_oi=34034784665600&from=timeline&isappinstalled=0

https://blog.csdn.net/jeryjeryjery/article/details/80245924

https://github.com/letiantian/TextRank4ZH

https://zhuanlan.zhihu.com/p/55316116