face_recognition | Rust bindings to dlibs face recognition tools | Computer Vision library

Modules in Python are not supposed to be run multiple times, they are supposed to be a way of defining variables to be used in other scripts. Therefore importing a module twice does not run it twice; the interpreter stores the variables from the module the first time it is run, and if you try to import it again it will give you the variables it already has stored and not run the file again.

There are a few ways of getting around this; here are two of them:

1. Put the code in the module into a function, and import that function:

The solution was as simple as possible:

We can use face_recognition.face_landmarks function instead to get landmarks.

Unfortunately I cannot test your code, but to me it seems like you are operation on a python list of tuples. You can sort that by using a key:

Fortunately, I solved the problem. These are the steps to download face_recognition Library:

in the command prompt write python and click Enter, this will open Microsoft store for you, "click download" or "install" to download and install python. "to check if it is downloaded or not type python in the command prompt

go to: https://pypi.org/project/pip-download/ click download files. Now open the downloaded file, then click right click with the mouse and choose "open in windows terminal". now type the command python get-pip.py. now your pip should be installed, type pip in the terminal to check if it is installed

go to: https://visualstudio.microsoft.com/ and download the community version of VS. after downloading install it and choose the following options in the installation page:
image1 image2

and choose install

Enter the command pip install cmake and wait fo it to be installed.

follow the instructions in the following link to know how to add something to your Path variable: https://helpdeskgeek.com/windows-10/add-windows-path-environment-variable/

write the following command in the command prompt: pip install dlib

write the following command in the command prompt: pip install face_recognition

click File click Sittings in the search type "interpreter" choose "python interpreter" choose "show all" click the "+" sign choose system Interpreter choose the interpreter that its path starts with the "users" directory

That is the solution I used to fix my problem

After some more research, I can up with this suggestion from Lewis Carroll, in this post "Beware the oom-killer, my son! The jaws that bite, the claws that catch!", and this post Chaining Chords produces enormously big messages causing OOM on workers, and this post WorkerLostError: Worker exited prematurely: exitcode 155.

It seems my celery workers may have been running out of memory, as I did find traces of the dreaded oomkiller in my syslogs. I reconfigured my tasks to just be in a chain (removed all the groups and chords) so each task is run individually in sequence for each image, and the tasks all completed successfully, no matter how many images I processed.

Mediapipe doesn't provide a face recognition method, only face detector.

The face_recognition library has really good accuracy, It's claimed accuracy is 99%+. your dataset probably isn't good enough.

Solutions:

1. For better speed performance, use the "hog" model instead of "cnn" model. you can modify it when you use the face_locations method like the following code line.

The container base python:3.9-slim is very stripped down. If your application requires CMake which often implies the gcc compiler as well, you have at least two options:

1. Use a more feature rich base container such as debian:buster
2. Choose a container with those tools already configured.

Example Dockerfile to build a base container:

Seems like you are using an outdated version of the dataset. You see the updated link here. The code you have will not work because you are asking for splits on data and target, while you are only providing target.

Something like this should work (on sklearn 0.24.1) :