This page helps you set up Python with MediaPipe and OSC capacities for interactive systems based on computer vision. Together, these components allow a wide range of applications. A proper setup is necessary to avoid conflicts.

Python Version

MediaPipe works with Python versions 3.8–3.12. To get the installed Python version, type the following into your terminal:

$ python3 --version

The output will provide the installed version (e.g. Python 3.12.3)

Getting the Right Python Version

If your Python is too old (<3.8) or too new (≥3.13), install or switch to a compatible version. Recommended versions: 3.10 - 3.12 (stable and widely supported). The following sections show the steps for the most common operating systems.

Windows

Uninstall older versions if needed from Add or Remove Programs.
Download the 64-bit installer from: https://www.python.org/downloads/windows/
During installation: - Tick Add Python to PATH - Under Customize installation, ensure pip and venv are selected.
Open a new PowerShell or Command Prompt and verify:
```
$ python --version
```

macOS

Option 1 – Use the official universal installer:

open https://www.python.org/downloads/macos/

After installation, confirm:

$ python3 --version

Option 2 – Use Homebrew (if installed):

$ brew install python@3.11
$ brew link python@3.11 --force --overwrite

Linux

On most systems you can install a supported version via package manager.

Ubuntu / Debian:

$ sudo apt update
$ sudo apt install python3.12 python3.12-venv python3.12-distutils

Then explicitly use that version:

$ python3.12 --version

Fedora:

$ sudo dnf install python3.11

Arch / Manjaro:

$ sudo pacman -S python

After installation, confirm the version again:

$ python3 --version

Check for pip

pip is Python's package installer - we need it to install dependencies. Make sure pip is available on your system:

$ python3 -m ensurepip --default-pip

Virtual Environment and Dependencies

We are using virtual environments to make sure we are using only the packages and versions we need for a specific project. This avoids confusions with different installs.

First, verify that pip is available, by typing the following in the terminal:

$ python3 -m venv --help

Create and activate a virtual environment - this step is the same for all operating systems:

$ python3 -m venv cv2pd

To activate te venv on macOS/Linux, enter:

$ source cv2pd/bin/activate

For Windows (PowerShell) type:

cv2pd\Scripts\activate

You will see that your venv is activated when the terminal shows (cv2pd).

Install Packages with pip

$ pip install mediapipe opencv-python python-osc numpy

You can verify the install by typing:

$ python -m pip list

You should see: opencv-python, numpy, python-osc