The tensorflow+imageai environment is built to image prediction and image detection

download

1. Installation and configuration of Anaconda and Pycharm https://www.cnblogs.com/yuxuefeng/articles/9235431.html

2. Conda update -all exception

exception handling

https://blog.csdn.net/qq_40100127/article/details/102608386

3. tensorflow and ImageAI ( It is recommended to use Terminal to install under the established project, so that this project needs to be used to prevent version inconsistencies between different projects from causing repeated problems )

Please refer to the URL first

https://www.sohu.com/a/254332853_100095143

tensorflow install

pip3 --default-timeout=100 install -i http://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com tensorflow==2.3.0

Or refer to:

Win10 system installs Tensorflow2.1-CPU version

https://blog.csdn.net/loft_/article/details/105550596

University of Science and Technology of China  https://pypi.mirrors.ustc.edu.cn/simple/
Tsinghua University  https://pypi.tuna.tsinghua.edu.cn/simple/
Alibaba Cloud  http://mirrors.aliyun.com/pypi/simple/
University of Science and Technology of China  http://pypi.mirrors.ustc.edu.cn/simple/
Douban  http://pypi.douban.com/simple/

imageai install

pip3 install https://github.com/OlafenwaMoses/ImageAI/releases/download/2.0.2/imageai-2.0.2-py3-none-any.whl

Or you can download imageai-2.0.2-py3-none-any.whl  , then run:

pip3 install C:\User\MyUser\Downloads\imageai-2.0.2-py3-none-any.whl

 

4. Bug handling of the running code after the environment is deployed

After installing Tensorflow, the specified module cannot be found

https://blog.csdn.net/loft_/article/details/105550867

 

ImageAI has an error of cannot import name 'convert_all_kernels_in_model' from 'tensorflow.python.keras.utils'

https://blog.csdn.net/Harry_Jack/article/details/105219586

 

 

Solve tensorflow error: AttributeError: module 'tensorflow.keras.backend' has no attribute 'get_session' problem

https://blog.csdn.net/weixin_41010198/article/details/107659012

 

TypeError: type object got multiple values for keyword argument 'training'

https://www.pythonf.cn/read/126614

 

 

tensorflow._api.v2.image' has no attribute 'resize_images'

https://github.com/fizyr/keras-retinanet/issues/1368

https://github.com/fizyr/keras-retinanet/issues/1368

5. Other examples collected

ImageAI implements a complete process: data set construction, model training, recognition and prediction

http://blog.itpub.net/69946223/viewspace-2655245

ImageAI Chinese document, the code can be used directly

https://imageai-cn.readthedocs.io/zh_CN/latest/ImageAI_Object_Detection.html

 

 

Guess you like

Origin blog.csdn.net/zkcharge/article/details/109258991