Skip to content
No description, website, or topics provided.
Jupyter Notebook
Branch: master
Clone or download
Latest commit 8bd97ae Aug 20, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Visualizing convolutional input Add files via upload Jul 30, 2018
Visualizing saliency maps Add files via upload Jul 30, 2018
README.md
Visualizing+image+classification+models+using+Saliency+Maps+and+Optimizing+Image.ipynb Add files via upload Aug 20, 2018

README.md

Visualising Image Classification Models and Saliency Maps

This project is based on the paper -Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps by Karen Simonyan, Andrea Vedaldi, Andrew Zisserman Visual Geometry Group, University of Oxford.(https://arxiv.org/pdf/1312.6034.pdf)

Here is a Keras implementation of above paper from scratch in the single notebook provided here.

The paper and project is divided into two parts:-

  1. Visualizing convolutional model input
  2. Visualizing saliency maps

Visualizing convolutional model input

Here, we will see the technique for visualising the class models learnt by the image classification ConvNets. Given a learnt classification ConvNet and a class of interest, the visualisation method consists in numerically generating an image, which is representative of the class in terms of the ConvNet class scoring model. We are going to use VGG16 pretrained network or carrying out this technique. This paper lets us know what are the features expected in input image to maximize the output node score.

What an ostrich looks like -

image

Visualization expected by output node-

ostrich

Visualizing Saliency maps

In saliency maps, we calculate the derivative of output of softmax class node with respect to the input image. Here we know the category of image beforehand. We just need to find out how does the output score changes with respect to each input pixel.This gives us an understanding of which parts of image are most discriminative.

Original image-

original

Saliency map-

saliency_map

Saliency map after smoothing-

saliency_map

You can’t perform that action at this time.