Skip to content

marlenezw/face-and-body-detector-with-mediapipe

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 

Repository files navigation

face-and-body-detector-with-mediapipe

This code detects face and body landmarks using mediapipe, a python ML package. (This code is adapted from nicknochnack's longer tutorial on using mediapipe for body language detection that can be found here.

It's important to note that MediaPipe Python on PyPI officially supports the 64-bit version of Python 3.7 to 3.10 on the following OS:

  • x86_64 Linux
  • x86_64 macOS 10.15+
  • amd64 Windows

You'll need to be using a virtual environment that's running one of these versions of Python.

Mac0S

(The following steps are courtesy of user josiahsrc on GitHub and stackoverflow

Here are the steps you need to take in order to use mediapipe with Apple's M1:

  1. Launch terminal using the Rosetta 2 translation layer. You can do this by opening Finder, going to Applications > Utilities and right clicking Terminal. In the right-click menu, click on get-info and then tick the Open Using Rosetta checkbox.

  2. Open a new Terminal window. (If terminal was previously opened, quit and relaunch it).

  3. Use the following command to install Homebrew for x86_64 architecture. arch -x86_64 /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install.sh)". We do this because we'd like to install Python 3.8 for x86_64 where mediapipe is supported.

  4. Now that we have brew for x86_64, we can install Python 3.8 for x86_64 by running the command arch -x86_64 /usr/local/homebrew/bin/brew install python@3.8. This installation may take a while.

  5. Now that we have the Python version we need, we can create a new Virtual Environment. By using the command arch -x86_64 /usr/local/homebrew/opt/python@3.8/bin/python3 -m venv myvenv. Remember, you need to specify your path to the x86_64 Python (the one we just installed). For me it was installed to /usr/local/homebrew/opt/python@3.8/bin/python3. For you it could be different. Either ways after Brew finishes installing Python, it'll display the path it was installed to. You should see "Python was installed at /Path/to/newly/installed/python". Also, the Virtual Environment will be created at ~/myvenv. You can change that by changing the myvenv part in the command.

  6. Start your Virtual Environment source ~/myvenv/bin/activate. I used this path ~/myvenv because thats where I created my Virtual Environment. If you created your Virtual Environment elsewhere, use that path.

  7. Now you should be inside the Virtual Environment, upgrade pip. pip install --upgrade pip

  8. Install mediapipe from pip, pip install mediapipe

That's it :) Now you can open your venv. Specify the interpreter as the path of the x86_64 Python we just installed (Example /usr/local/homebrew/opt/python@3.8/bin/python3). Now you should be able to use mediapipe.

example output

About

This code detects face and body landmarks using mediapipe.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published