grad-cam.tensorflow

by Ankush96

Implementation of Grad CAM in tensorflow

210 Stars 73 Forks Last release: Not found 7 Commits 0 Releases

Available items

No Items, yet!

The developer of this repository has not created any items for sale yet. Need a bug fixed? Help with integration? A different license? Create a request here:

grad-cam.tensorflow

Implementation of Grad CAM in tensorflow

Gradient class activation maps are a visualization technique for deep learning networks.

The original paper: https://arxiv.org/pdf/1610.02391v1.pdf

The original torch implementation: https://github.com/ramprs/grad-cam

Setup

Clone the repository

sh
git clone https://github.com/Ankush96/grad-cam.tensorflow/
Download the VGG16 weights from https://www.cs.toronto.edu/~frossard/vgg16/vgg16_weights.npz

Usage

python main.py --input laska.png --output laska_save.png --layer_name pool5

Results

| Input | Output | | ------|-----:| | Original image | Original image + Visualization |

Acknowledgement

Model weights (vgg16weights.npz), Class names (imagenetclasses.py) and example input (laska.png) were copied from this blog by Davi Frossard (https://www.cs.toronto.edu/~frossard/post/vgg16/). TensorFlow model of vgg (vgg16.py) was taken from the same blog but was modified a little. https://github.com/jacobgil/keras-grad-cam also provided key insights into understanding the algorithm.

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.