/DL-CNN

Implementation of a Convolutional Neural Network

Primary LanguagePythonMIT LicenseMIT

DL-CNN

Implementation of a Convolutional Neural Network using Python

About

This is the implementation of a CNN on MNIST data set. It's basically a handwritten digit database of 60000 training example and 10000 testing examples. It is recognition of an image on which a digit is written in a normal handwriting.

Convolution Neural Network is a class of Deep Neural Networks which is mainly deployed for Computer Vision tasks. It is arguably the most widely used and popular architecture in deep learning. The speciality of CNN’s is that the network requires minimum pre-processing compared to other network architectures and algorithms. It mainly works on data which has a grid pattern – Image Data and it is designed in a way to learn the spatial set of features and patterns.

Applications

  1. Face Recognition
  2. Image Classification
  3. Action Recognition
  4. Human Pose Estimation
  5. Document Analysis

Pre-requisites

Download and install the following module:
1)tensorflow- pip install tensorflow

Author

Pranav Raikote