mihreteab-H/tigrigna-handwritten-character-recognition
Domain:
natural language processingRecord type:
model
Creator:
mih
Host:
AI-based Tigrigna handwritten character recognition using CNN and TensorFlow
# AI-Based Tigrigna Handwritten Character Recognition
## Overview
This project implements a Convolutional Neural Network (CNN) to recognize Tigrigna handwritten characters.
## Features
- Image preprocessing (grayscale conversion, resizing, normalization)
- Data augmentation
- CNN model training and evaluation
- TensorFlow implementation
## Tech Stack
- Python
- TensorFlow / Keras
- NumPy
- OpenCV
- Google Colab
## Results
Achieved high validation accuracy on the dataset for character classification.
## Future Improvements
- Deploy model using FastAPI
- Docker containerization
- Build web interface for predictions
## Model Performance
Validation Accuracy: 92%
Similar
sohankandagatla/Handwritten-Amharic-Character-Recognitionnatenaile/Handwritten-Amharic-Character-RecognitionshraddhaA2/handwritten-amharic-character-recognitionLearning based Ge'ez character handwritten recognitionFetulhak/Handwritten-Amharic-Character-Recognition-systemEsubalew-kunta/Amharic-Handwritten-character-recognition-system