/tensorflowvgg19

Primary LanguagePython

Visualization of Deep Covolutional Neural Networks

Build Status Coverage Status

  • This repository contains implementations of visualizatin of CNN in recent papers.
  • The source code in the repository can be used to demostrate the algorithms as well as test on your own data.

Requirements

Algorithms

Gradient-weighted Class Activation Mapping (Grad-CAM)

  • Grad-CAM generates similar class heatmap as CAM, but it does not require to re-train the model for visualizatin.
  • Details of the implementation and more results can be find here. Some results:

grad-cam-result

Class Activation Mapping (CAM)

  • The class activation map highlights the most informative image regions relevant to the predicted class. This map can be obtained by adding a global average pooling layer at the end of convolutional layers.
  • Details of the implementation and more results can be find here. Some results:

celtech_change

Guided back propagation

  • Details of the implementation and more results can be find here. Some results:

gbp