CNN-alphabet
v4han/CNN-alphabet
Model trained with Armenian alphabet using CNN 95% accuracy
computer-visioncharacter-recognitionconvolutional-neural-networkimage-classificationtensorflowkaggle-competitionarmenian-language
Summary
A Convolutional Neural Network (CNN) model for recognizing characters of the Armenian alphabet, achieving ~95% accuracy. The project is based on a Kaggle competition dataset (Mashtots Dataset) and includes a Jupyter notebook for training, a saved model, and basic setup instructions.