Classify 102 Flowers

This model was created by science4performance, using the fast.ai library.
It was trained using the dataset created and kindly made available by Maria-Elena Nilsback and Andrew Zisserman,
Visual Geometry Group, Department of Engineering Science University of Oxford, United Kingdom
The model achieved an accuracy of 80% on a test set of 102 flowers. Improved version achieved 97% accuracy.
It works best on full frame close up images of a single flower

Chosen Image