目录
1、模型准备工作
1-1 模型运行环境
python3.6
TensorFlow1.12-GPU
windows10
Ananonda
Pycharm
CUDA9+cuDNN7.6
opencv-python
tensorflow-plot
cython
easydict
Pillow
matplotlib
scipy
1-2 模型地址
https://github.com/dBeker/Faster-RCNN-TensorFlow-Python3
1-3 数据集及网络地址
VOC2007：https://uinedu-my.sharepoint.com/:f:/g/personal/19604_myoffice_site/EiLTzAbNirROrQQF20eupMQB-KpIfZOa7w2YS5MB2ARvSA
http://download.tensorflow.org/models/vgg_16_2016_08_28.tar.gz
resnet50_v1:http://download.tensorflow.org/models/resnet_v1_50_2016_08_28.tar.gz
resnet101_v1:http://download.tensorflow.org/models/resnet_v1_101_2016_08_28.tar.gz
2、模型训练
2-1 train前准备
打开cmd，激活TensorFlow-GPU环境，并转入模型中的下述路径
D:\pytrain\my_models\Faster-RCNN-TensorFlow-Python3-master\data\coco\PythonAPI
运行命令
python setup.py build_ext --inplace
python setup.py build_ext --install
然后转到模型文件中的此路径
[En]
Then go to this path in the model file
D:\pytrain\my_models\Faster-RCNN-TensorFlow-Python3-master\lib\utils
运行命令
python setup.py build_ext --inplace
当运行上述三个命令时，您将在此文件夹下的几个红色框中找到这些文件。
[En]
When the above three commands are run, you will find the files in several red boxes under this folder.
D:\pytrain\my_models\Faster-RCNN-TensorFlow-Python3-master\data\coco\PythonAPI\pycocotools
此文件夹中将有更多红盒文件。
[En]
There will be more red box files in this folder.
D:\pytrain\my_models\Faster-RCNN-TensorFlow-Python3-master\lib\utils
随后第一部分的链接中下载公开VOC数据集，将其放在下述路径
D:\pytrain\my_models\Faster-RCNN-TensorFlow-Python3-master\data\VOCdevkit2007\VOC2007
2-2 下载vgg16及resnet模型
将下载的模型放入以下文件夹
[En]
Put the downloaded model into the following folder
D:\pytrain\my_models\Faster-RCNN-TensorFlow-Python3-master\data\imagenet_weights
其中imagenet_weights文件夹需要自己创建
2-3 train
运行train.py即可进行训练
Original: https://blog.csdn.net/weixin_49764531/article/details/123633147
Author: Be_Better_World
Title: windows10下TensorFlow版本Faster RCNN复现
原创文章受到原创版权保护。转载请注明出处：https://www.johngo689.com/509145/
转载文章受原作者版权保护。转载请注明原作者出处！