강아지 상태를 추정해주는 딥러닝 모델
이 코드는 밑바닥부터 시작하는 딥러닝 책을 보면서 작성하였습니다.

I wrote this code while looking at the book 'Deep Learning from Scratch'.

Python3 and later
numpy
opencv

clone https://github.com/WikiCloggy/cloggy_extractor install dlib install yolo with https://github.com/thtrieu/darkflow mkdir weights download wieghts from https://github.com/thtrieu/darkflow

Class cloggyNet is the neural network designed for trainig the cloggy states with skeleton silhouette data.
Class cloggy_state_estimator is the trained cloggyNet and it can predict a cloggy state.
If you look at the demo.py, you can get information about how to predict cloggy state with cloggy_state_estimator.

python start_estimate.py test/image/test1.jpg

The result will be saved into data/result with jason format
The input image will be saved into data/input_image The input image will be changed to silhouette for estimation and saved into data/input_silhouette

python start_estimate.py test/image/test1.jpg -flip False

If you get more state skeleton data, than just run cloggyNet_trainer.py.
If you want to add more state, add more state to label in label_maker.py and run. Then, add data to created label directories.

If you want to make skeleton silhouette data, than use cloggy_dataset_maker in https://github.com/WikiCloggy/cloggy_dataset_maker!