Skip to content

Latest commit

 

History

History
20 lines (15 loc) · 686 Bytes

README.md

File metadata and controls

20 lines (15 loc) · 686 Bytes
Raw

Visualizing Convolutional Neural Network (CNN)

This project aims to visualize the layers of CNN using MNIST digit dataset.

The layers of the network are as follows :

  • Input Layer
  • Convolutional Layer ( 8 Filters )
  • Pooling Layer
  • Convolutional Layer ( 16 Filters )
  • Pooling Layer
  • Convolutional Layer ( 32 Filters )
  • Pooling Layer
  • Dense Layer ( 100 Units )
  • Dense Layer ( 10 Units )
  • Output Layer

The model has been trained using Tensorflow ( Python3 ) and loaded into Tensorspace for visualization. The user input is fed to the model using Canvas as input.

A better training on the network is required. The main aim of doing this project was visualization.