ramapinnimty/Hindi-Character-Classifier

A multi-level (2 levels) CNN model was built using Keras to classify Hindi characters from 5 classes. The training and validation accuracies were 99.2% and 98.9% respectively. The model was then used to predict the class labels of new images and it has done its job quite well. All the results and statistics are clearly mentioned in the notebook along with interactive visualizations.

Jupyter NotebookMIT