Free tool to create your custom HUMAN POSE ANNOTATION data sets for Machine Learning/Computer Vision Applications!
git clone this repo to your local machine
git clone https://github.com/moonbeam5115/PoseAnnotation.git
cd PoseAnnotation
Set up virtual environment:
(Conda)
conda create --prefix ./env python=3.7
Activate conda environment:
conda activate ./env
Install requirements:
pip3 install -r requirements.txt
Place images you wish to annotate inside the imgs folder
Start program:
python main.py
Left Click on image to add joint annotation
Press "e" to undo previous joint annotation
Press "enter" to save results and go to next image -- Results are saved to the RESULTS Directory
Press "escape" to exit program
Create functionality to return to previous image
Allow user to skip occluded joints
If this project helped your research or other work in any way, consider donating to my Patreon to support other ongoing Computer Vision/Machine Learning/Data Science projects.
Head over to:
Patreon.com/supermanecuavoley