Weather Image Recognition is a project which aims to create a machine learning model based on Convolutional Neural Network with an objective of recognising weather present in images. The dataset is characterized by a high diversity of images.
Train accuracy: 74.87%
Test accuracy: 68.82%
π¦WeatherImageRecognition
β£ πdataset
β β£ πtest
β β β£ πdew
β β β β£ π2220.jpg
β β β β£ π2225.jpg
β β β β ...
β β β£ πfogsmog
β β β β£ π4087.jpg
β β β β£ π4092.jpg
β β β β ...
β β β£ πfrost
β β β β£ π3609.jpg
β β β β£ π3633.jpg
β β β β ...
β β β£ πglaze
β β β β£ π6102.jpg
β β β β£ π6107.jpg
β β β β ...
β β β£ πhail
β β β β£ π0012.jpg
β β β β£ π0017.jpg
β β β β ...
β β β£ πlightning
β β β β£ π1839.jpg
β β β β£ π1863.jpg
β β β β ...
β β β£ πrain
β β β β£ π1044.jpg
β β β β£ π1075.jpg
β β β β ...
β β β£ πrainbow
β β β β£ π0619.jpg
β β β β£ π0634.jpg
β β β β ...
β β β£ πrime
β β β β£ π4955.jpg
β β β β£ π4965.jpg
β β β β ...
β β β£ πsandstorm
β β β β£ π2920.jpg
β β β β£ π2925.jpg
β β β β ...
β β β πsnow
β β β β£ π0842.jpg
β β β β£ π0847.jpg
β β β β ...
β β£ πtrain
β β β ...
β β£ πvalid
β β β ...
β£ π.gitignore
β£ πREADME.md
β πWeather_image_recognition.ipynb
- TensorFlow
- Keras
- Python