Here you can easily generate the trainging dataset for YOLOv5 in the following steps...
conda create --name eyestrackingyolov5 python=3.7
conda activate eyestrackingyolov5
git clone https://github.com/boyleerock/Eyes-Tracking-s-Dataset-Generator
cd Eyes-Tracking-s-Dataset-Generator
pip install -r requirements.txt
-
Download the pretrained model shape_predictor_68_face_landmarks.dat
-
Then CHANGE your images folder's name in loop_directory("images/") in findeyes_convert2yoloformat .py
python findeyes_convert2yoloformat.py
