In this project, we discuss on how to correctly get valid pose and face estimations when we pass images through YOLO-v8 model and MediaPipe Face detection model respectively, and we will see how well the pretrained models actually perform, which makes them so versatile and easy to use.
Moreover, the accuaracy of the pretrained models is so high that practically no fine-tuning is required.
We have attached the relevant dataset in the google drive link below.
https://drive.google.com/drive/folders/1KM7YWeeqyuUMlO54cNhLZo6L8yWP_dXa?usp=drive_link