GitHub - rozana107/Simple-CNN-classification: A PyTorch-based CNN project for classifying weather images into four categories: cloudy, rain, shine, and sunrise. The code includes image preprocessing, dataset cleaning, train/test splitting, CNN model training, evaluation, and prediction, achieving around 96% validation accuracy on a small weather image dataset.

#Simple-CNN-classification

Unzip the folder resized.zip befor running the code.

The original data files exceed the limit to be uploaded.

You can modify the code and use the resized images.

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
README.md		README.md
classification-64.ipynb		classification-64.ipynb
classification-64.py		classification-64.py
classification-conf_mat.ipynb		classification-conf_mat.ipynb
classification.pdf		classification.pdf
resized.zip		resized.zip
weather_net_30_0_96.pth		weather_net_30_0_96.pth
weather_net_55_0_95.pth		weather_net_55_0_95.pth

Navigation Menu