Introduction to facial expression recognition system -- offline environment configuration

Abstract ： This article details how to use the offline method to configure the previous blog post 《 Introduction to facial expression recognition system —— Part 1 （python Realization , contain UI Interface and complete code ）》 Code in project Python Environmental Science , It can effectively avoid some possible dependent package installation problems . The software required for this offline package includes PyCharm and Anaconda, Bloggers have packed and shared through Baidu online disk , So that the bloggers who pay attention to the project can download, install and use , The main points of the blog are as follows ：

---

Preface

         Blog posts written by previous bloggers Introduction to facial expression recognition system —— Part 1 （python Realization , contain UI Interface and complete code ） Shared an expression recognition project , Many fans have sent messages asking about the environment configuration method . Many new friends can't configure Python Environmental Science , Or it is difficult to install some dependent packages due to the computer network environment , So here's what the blogger packed Python Offline environment , For everyone to transplant to their own computers , In order to solve the problem of inconsistent environment versions .

---

1. Download code and offline environment

1.1 Download code

        （1） First of all, of course, is to download the code for facial expression recognition , The download address is as follows , As shown in the figure, we click on the page 2.1 Download link to version , Start downloading and extracting the files ：

Download code ：https://mianbaoduo.com/o/bread/YpialJ0=

        （2） The downloaded and unzipped folder is shown in the figure , Please put it in a familiar folder path without Chinese ：

1.2 Download the dependent package file

        （1） The blogger is ready to run all the code required for this project Python Dependency package , This step requires you to download the ready offline Python Environment package .

         link ：https://pan.baidu.com/s/1NJS7ufGkPw_b2lor2lockw
         Extraction code ：fxrd

        （2） First, click the above link to download the offline dependency package , And will download “Pypackage” Folder on top “Emotion Recognition” In the folder , That is, the directory looks like the one shown in the figure below ：

         Download and place this folder , In this way, the program code and dependent package for facial expression recognition are ready . The next step is just to PyCharm Create a new one in Conda Environment and install these dependencies , Last in PyCharm You can run this program on your own computer by specifying this environment in .

---

2.PyCharm And Anaconda install

         Before we configure the environment and run the code ,PyCharm and Anaconda It must be installed . Even though PyCharm It can also be installed and managed Python Of dependent packages , However, in order to facilitate the configuration and maintenance of offline packages in the future , Please install Anaconda This must-have artifact . The installation steps for these two software are basically one click Next, Simple installation , There are many tutorials on the Internet , You can refer to the installation . Of course, I recommend using the installation tutorial written by the blogger ：PyCharm And Anaconda Super detailed installation and configuration tutorial , Each step of the installation process is described in detail . Just click the link and install it , The software installation will not be repeated here .

---

3. Install the offline dependency package

3.1 newly build Conda Environmental Science

        （1） First let's open up PyCharm, Click... In the upper left menu bar “File”, choice “Open”, In the pop-up browse option box, select the folder directory extracted in the previous section （ That is, the folder of our project code ）, Click on “OK” You can open the project , As shown in the figure below ：

        （2） The project interface after opening is shown in the following figure , Click on the bottom left “Terminal” May open PyCharm It comes with terminal tools . Its function and CMD The terminal is the same , It's just that here when we open the project ,PyCharm The directory of the terminal is automatically switched to the folder path where the project is located , It saves us the steps of switching paths .

        （3） Let's first create a new one for this project Conda Environmental Science , Enter the following code in the terminal , Create a new one called "emo_rec" Environment , And designate Python The version is 3.7（ For this project, please use 3.7 Version of ）, As shown in the figure below ：

conda create --name emo_rec python=3.7

        （4） Press enter , Start creating environment , We will be prompted to confirm the installation , type “y” And return , You can start the installation ：

        （5） At this time, you can see the following interface without any accident , Show that the environment configuration is complete , And prompt us to activate the environment ：

        （6） Let's turn off this terminal first , Then restart the terminal , Enter the following code to activate the environment we just created “emo_rec”, And go back to execute ：

conda activate emo_rec

        （7） completion of enforcement , You can see that the header of the command line has been displayed as “(emo_rec)”, Indicates that you are currently in emo_rec In the environment ：

3.2 Start installing dependency packages

        （1） You are now ready to install the offline dependency package . First open and find “Pypackage” In folder “setup.txt” file , The installed packages and corresponding versions have been listed in it , Here you need to copy it to “Emotion Recognition” Under the folder .

         To ensure that the following steps are carried out correctly, several points need to be confirmed ： First of all ,“Pypackage” The folder and “setup.txt” File in Unzipped “Emotion Recognition” Under the folder ; second , The terminal path is in “Emotion Recognitive” Under the path , Third , The terminal is already in the new “emo_rec” In the environment , It is roughly as shown in the figure below ：

        （2） So we just need to use pip install setup.txt The dependent packages listed in the file are OK , Enter the following code in the current terminal and press enter , Officially start the offline package installation .

pip install -r setup.txt --no-index --find-links=./Pypackage/

         The above code means to use pip install setup.txt The packages listed in the file , It also specifies that the directory of the offline installation package is in the... Of the current directory Pypackage Under the folder , As shown in the figure below ：

        （3） Type the code and press enter to wait for the dependent package to be installed , The general display of successful installation is as follows ：

         So far, we have created and named our own environment by relying on packages offline "emo_rec" The required third-party dependency libraries are installed . Readers can see that the previous steps are basically consistent with our usual configuration environment , Just steps （2） There are conditions to specify the next good installation package , This actually avoids many problems that may occur during online installation .

3.3 stay PyCharm Configure the environment in

        （1） Start configuring the environment , Now we just need to PyCharm Specify the newly installed environment as the running environment of the facial expression recognition project . Click on “File” menu , choice “Setting” Options , stay “Project” Find “Python Interpreter”, If it has not been configured before, the environment display should still be empty ;

        （2） Click the environment drop-down box （ The picture below is red 2 It's about ）, Click... Below the drop-down box “Show All…”, The environment display interface pops up , Click on the bottom left corner “ plus ” Icon , Get into Python Environment selection interface , Choose “Conda Environment”, Choose “Existing environment”（ As shown in Figure No 6 Step ）, At this time, you can choose 3.2 Section conda Environmental Science , The address is “anaconda The installation directory \envs\ Your environment name \python.exe”, Finally, click “OK”;

         Normally, we just installed conda Environmental Science , Above, red 7 The office will automatically help us choose this environment , If it doesn't appear, it can be found in your anaconda Installation directory conda Folder , One of the “envs” Folder , You can find the folder of environment name in it , Choose python.exe That's it .

        （3） Click on OK after , At this point, the installed dependency packages can be seen on the environment interface , Here's the picture , Click on “OK” that will do

        （4） As a result of later updates Keras The... In this bag saving.py File is incompatible , So we need to download the “Pypackage” In folder “saving.py” File copy to “ Yours anaconda The installation directory \conda\envs\emo_rec\Lib\site-packages\keras\engine\” Under the folder （emo_rec It's the environment name ）, Replace the original “saving.py” file ：

        （5） After some operation , We can open... In the file bar on the left "runMain.py" File code , And right-click in the editor , choice “Run runMain”, You can run the code

        （6） At this point, the program can run successfully , The interface is shown in the following illustration ：

---

Conclusion

         Because of the limited ability of bloggers , The methods mentioned in the blog post are even tested , There will inevitably be omissions . I hope you can enthusiastically point out the mistakes , So that the next modification can be more perfect and rigorous , In front of you .Python There are also many online tutorials for software installation , The installation steps need not be the same , Due to many factors , Different people may install it differently , Therefore, if the installation fails, it is normal , Keep calm , Problem solving is the most important thing , At the same time, if there is a better implementation method, please don't hesitate to give us your advice .