Installation#
This section explains how to install the required software stack natively on your system. However, you might prefer using CompressAI-Vision via docker instead.
1. Python dependencies#
Here we assume you are running Ubuntu 20.04 LTS (or newer)
First, create and activate the virtualenv with:
python3 -m venv venv
source ./venv/bin/activate
pip install -U pip
You might want to define the python version explicitly, i.e. with python3.8 -m venv venv
. Python3.8+ is required.
While in the activated virtualenv, run the installation:
bash scripts/install.sh
The script will install the following software stack:
This library (CompressAI-Vision)
After running the script within the virtualenv, you might need to deactivate
and activate
it for the effects to take place.
2. If your pipeline includes traditional codecs#
Install binaries from sources, clone the selected repo from the following links and select the desired tag
VVENC: <fraunhoferhhi/vvenc>
Please refer to the respective Readmes to build the binaries, that will be launched from compressai-vision code.