Skip to content
Hindi Alphabet classifier using convnet
Python
Branch: master
Clone or download
Latest commit a7baab1 Aug 22, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Dev-Rec.py initial commit Apr 28, 2018
Hindi_Letters.py Update Hindi_Letters.py Apr 29, 2018
LICENSE.txt initial commit Apr 28, 2018
data.7z adding data Apr 28, 2018
devanagari_model.h5 initial commit Apr 28, 2018
hindi.gif adding gif Apr 28, 2018
hindi_digits.csv added csv with hindi digits Apr 29, 2018
readme.md Update readme.md Aug 22, 2018
requirements.txt initial commit Apr 28, 2018

readme.md

Devanagiri Recognition

This code helps you classify different alphabets of hindi language (Devanagiri) using Convnets

Sourcerer

Code Requirements

You can install Conda for python which resolves all the dependencies for machine learning.

Description

This code successfully recognizes hindi characters.

Technique Used

I have used convolutional neural networks. I am using Tensorflow as the framework and Keras API for providing a high level of abstraction.

Architecture

CONV2D --> MAXPOOL --> CONV2D --> MAXPOOL -->FC -->Softmax--> Classification

Some additional points

  1. You can go for additional conv layers.
  2. Add regularization to prevent overfitting.
  3. You can add additional images to the training set for increasing the accuracy.

Python Implementation

  1. Dataset- DHCD (Devnagari Character Dataset)
  2. Images of size 32 X 32
  3. Convolutional Network Support added.

Train Acuracy ~ 95%

Test Acuracy ~ 92%

Execution for writing through webcam

To run the code, type python Dev-Rec.py

python Dev-Rec.py
You can’t perform that action at this time.