arabichar-balanced-ver3

Model Card for Model ID

This CNN fine-tuned model is designed to identify Alphabet characters written by children for ages between 4-8.

Model Details

Model Description

<!-- Provide a longer summary of what this model is. -->

This model focuses on adapting a pre-trained CNN model. It is built of three convolutional layers with 32 filters, followed by max-pooling and batch normalization. Another set of three convolutional layers with 64 filters extracts deeper features, followed by another pooling and normalization step. The extracted features are passed through two fully connected layers with dropout, and the final softmax layer classifies the characters into 28 categories.

The base model, trained on AHCD, and I've fine-tuned it on Dhad-Hijja Dataset collection,

• Developed by: [shahad-alh]
• Model type: [Image Classification]
• Language(s) (NLP): [ar (Arabic)]
• License: [MIT]
• Finetuned from model [optional]: [More Information Needed]

Model Sources [optional]

<!-- Provide the basic links for the model. -->

• Repository: [More Information Needed]

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

Could be used by anyone who is interested in buliding App, Model, anything that should identify Children's characters handwriting.

Evaluation

Training Accuracy :86%

Test Accuracy: 87%

🚀 Demo

Try it out live here: