dima806/flowers_16_types_image_detection

Returns flower type given an image with about 99.5% accuracy.

See https://www.kaggle.com/code/dima806/flowers-16-types-image-detection-vit for more details.

Classification report:

                  precision    recall  f1-score   support

       calendula     0.9928    0.9786    0.9856       421
       coreopsis     0.9882    0.9905    0.9893       421
            rose     0.9976    0.9953    0.9964       422
black_eyed_susan     0.9976    0.9976    0.9976       422
      water_lily     0.9953    1.0000    0.9976       421
california_poppy     0.9905    0.9929    0.9917       422
       dandelion     1.0000    0.9976    0.9988       422
        magnolia     0.9952    0.9858    0.9905       422
         astilbe     0.9976    0.9976    0.9976       421
       sunflower     0.9976    1.0000    0.9988       422
           tulip     0.9976    1.0000    0.9988       422
      bellflower     0.9952    0.9905    0.9929       422
            iris     1.0000    1.0000    1.0000       421
    common_daisy     0.9882    0.9952    0.9917       421
        daffodil     0.9976    0.9976    0.9976       422
       carnation     0.9859    0.9976    0.9918       422

        accuracy                         0.9948      6746
       macro avg     0.9948    0.9948    0.9948      6746
    weighted avg     0.9948    0.9948    0.9948      6746