RodolGiaco / SmartHealthyPostureDetector

Detector de postura inteligente utilizando redes neuronales y vision artificial
0 stars 0 forks source link

ThirdPerson

This repository contains the code developed for the undergraduate thesis project developed between September and December 2023 for the Polytechnic University of Madrid.

ThirdPerson lets users control a drone by using gestures. It has 2 modes: not following and following. The latter tracks the user with the information provided by MediaPipe.

Installation

To use ThirdPerson, you can either install it by using Docker or by running the following command:

pip3 install -r requirements.txt

Usage

Connect the Tello drone and execute main.py. If you do not have a drone, you can activate the simulation option in config.json, which will access your webcam.

With the app running, 3 windows should open, as seen in the picture below:

gui

Point to the drone to change the following mode.

Customization

ThirdPerson is designed to be highly customizable. Each module can be used independently to implement individual functions. To add gestures, run model/add_gesture.py. To retrain the neural network, run the cells in model/keypoint_model_training.ipynb.

Other stuff

To read about the design behind this project, see the thesis (in Spanish)