This pipeline produces a copy of the original video projecting the 3d pose estimation produced by the cv model MeTRAbs (https://doi.org/10.48550/arXiv.2007.07227).
If you are an Unix system based user, just follow these steps
$ git clone https://github.com/almeidadm/pose_estimation.git
$ cd pose_estimation
$ python -m venv .venv
$ source ./source/bin/activate
$ python -m pip install -r requirements.txt
first of all copy your reference video to data_input directory, then just execute
$ python main.py
The data_output folder contains: - original video - the original frames - the processed frames - the produced video