GitHub - Davidooff/AI-pill-detection

This code build with model VGG16 with changed Dense layers

Dense: 128, 64, 32 - relu

Out: 4 - sigm

U can run this code buy

python3 main.py [-option] [-option] [-option]

-l - load weights from ./weights.h5

-train - start train

-train-save - train and save weights as ./new-weights.h5

-train-save-show - same as up, but also showing

-p - predict from DataSet/test

Provide feedback

Name		Name	Last commit message	Last commit date
Latest commit History 16 Commits
DataSet		DataSet
__pycache__		__pycache__
.gitignore		.gitignore
README.md		README.md
config.py		config.py
gray_mse_1e-3_5ep_D-128-64-32.png		gray_mse_1e-3_5ep_D-128-64-32.png
gray_mse_1e-4_10ep_D-128-64-32.png		gray_mse_1e-4_10ep_D-128-64-32.png
main.py		main.py
rgb_mse_1e-4_10ep_D-128-64-32.png		rgb_mse_1e-4_10ep_D-128-64-32.png
rgb_mse_1e-4_10ep_D-256-128-64-32.png		rgb_mse_1e-4_10ep_D-256-128-64-32.png
rgb_mse_5e-5_10ep+10+5_D-128-64-32.png		rgb_mse_5e-5_10ep+10+5_D-128-64-32.png
rgb_mse_5e-5_10ep+15+5_D-128-64-32.png		rgb_mse_5e-5_10ep+15+5_D-128-64-32.png
rgb_mse_5e-5_10ep+20+10_D-128-64-32.png		rgb_mse_5e-5_10ep+20+10_D-128-64-32.png
rgb_mse_5e-5_10ep+30+10_D-128-64-32.png		rgb_mse_5e-5_10ep+30+10_D-128-64-32.png
rgb_mse_5e-5_10ep+40+10_D-128-64-32.png		rgb_mse_5e-5_10ep+40+10_D-128-64-32.png
rgb_mse_5e-5_10ep+5+5_D-128-64-32.png		rgb_mse_5e-5_10ep+5+5_D-128-64-32.png
rgb_mse_5e-5_10ep+50+10_D-128-64-32.png		rgb_mse_5e-5_10ep+50+10_D-128-64-32.png
rgb_mse_5e-5_10ep+5_D-128-64-32.png		rgb_mse_5e-5_10ep+5_D-128-64-32.png
rgb_mse_5e-5_10ep_D-128-64-32.png		rgb_mse_5e-5_10ep_D-128-64-32.png