Real-time-low-cost-multi-person-3D-pose-estimation

Contact : Alice Ruget [email protected]

This folder contains the code and models of the project.

You can find the models and the data at https://researchportal.hw.ac.uk/en/datasets/real-time-low-cost-multi-person-3d-pose-estimation.

We provide one example for each scenario containing one, two or three people in the scene.

Dependencies

Python 3.8.11 Tensorflow 2.4.1 Keras 2.4.0

Run the tests

1. Download the models at the DOI : 10.17861/e85a6eae-13f9-4bcd-9dff-73f8107c09a2
2. Run Pixels2Pose.py --scenario=number_people with number_people = 1,2,3

Results

1 people scenario

2 people scenario

3 people scenario