IndCurr

Resnet34 trained for Indian Currency Recognition

Inference Examples

Streamlit App

Accuracy of each class on test dataset

Classes : ['Rs 10','Rs 20', 'Rs 50', 'Rs 100', 'Rs 200', 'Rs 500', 'Rs 2000']
Class accuracy : [ 96.29, 87.87, 88.88, 90.909, 91.66, 100, 100]

Confusion matrix

Real world use cases:

Currency recognition system from visually impaired/blind peoples.
Scope for development of currency verification systems.