You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Amharic Handwritten Character Classifier

Overview

This project implements a deep learning model for recognizing Amharic handwritten characters. The dataset includes 238 unique Amharic characters covering different forms and variations.

The model predicts characters from 28x28 grayscale images and outputs both the predicted label and its confidence.

Dataset

Dataset collected from various sources covering different age ranges and writing orientations.
Split into train/val/test sets.
Each image filename encodes its label, corresponding to an Amharic character.

Preprocessing

Images normalized to [0, 1].
Organized into subfolders per class.
Labels saved in a JSON file mapping class indices to actual Amharic characters.

Model

Convolutional Neural Network (CNN) with Conv2D, MaxPooling, and Dense layers.
Output: 238 softmax units.
Trained with categorical_crossentropy loss and Adam optimizer.

Testing & Prediction

Model tested on random samples from the test set.
Predictions include actual Amharic label from the JSON file.
Each prediction shows the confidence level, derived from the softmax output.

Example output:

Sample:
  True Label: ሀ
  Predicted: ሀ
  Confidence: 0.9942

Usage

Upload an image and the model will predict the character with its actual Amharic label and confidence.

Notes

Model file: amharic_character_classifier.h5
Labels file: amharic_labels.json
Predictions map class indices to actual Amharic characters.

📌 Acknowledgements

Dataset: Fetulhak Abdurahman
Implementation & training: tamratD
Frameworks: TensorFlow, Keras, Google Colab

📬 Contact

For dataset-related questions, contact: afetulhak@yahoo.com. For implementation details: Hugging Face profile

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

tamratD
/

amharic-character-classifier