xiaoganghan/saliency

Implementation of SmoothGrad as well as several other saliency techniques

Saliency Methods

Introduction

This repository contains code for SmoothGrad, as well as implementations of several other saliency techniques. Each of these techniques can also be augmented with SmoothGrad. The techniques implemented in this library are:

• Vanilla Gradients (paper, paper)
• Guided Backpropogation (paper)
• Integrated Gradients (paper)
• Occlusion

This list is by no means comprehensive. We are accepting pull requests to add new methods!

Download

git clone https://github.com/tensorflow/saliency
cd saliency

Usage

Each saliency mask class extends from the SaliencyMask base class. This class contains the following methods:

• __init__(graph, session, y, x): Constructor of the SaliencyMask. This can modify the graph, or sometimes create a new graph. Often this will add nodes to the graph, so this shouldn't be called continuously. y is the output tensor to compute saliency masks with respect to, x is the input tensor with the outer most dimension being batch size.
• GetMask(x_value, feed_dict): Returns a mask of the shape of non-batched x_value given by the saliency technique.
• GetSmoothedMask(x_value, feed_dict): Returns a mask smoothed of the shape of non-batched x_value with the SmoothGrad technique.

The visualization module contains two visualization methods:

• VisualizeImageGrayscale(image_3d, percentile): Marginalizes across the absolute value of each channel to create a 2D single channel image, and clips the image at the given percentile of the distribution. This method returns a 2D tensor normalized between 0 to 1.
• VisualizeImageDiverging(image_3d, percentile): Marginalizes across the value of each channel to create a 2D single channel image, and clips the image at the given percentile of the distribution. This method returns a 2D tensor normalized between -1 to 1 where zero remains unchanged.

If the sign of the value given by the saliency mask is not important, then use VisualizeImageGrayscale, otherwise use VisualizeImageDiverging. See the SmoothGrad paper for more details on which visualization method to use.

Examples

This example iPython notebook shows these techniques is a good starting place.

Another example of using GuidedBackprop with SmoothGrad from TensorFlow:

from guided_backprop import GuidedBackprop
import visualization

...
# Tensorflow graph construction here.
y = logits[5]
x = tf.placeholder(...)
...

# Compute guided backprop.
# NOTE: This creates another graph that gets cached, try to avoid creating many
# of these.
guided_backprop_saliency = GuidedBackpropSaliency(graph, session, y, x)

...
# Load data.
image = GetImagePNG(...)
...

smoothgrad_guided_backprop =
    guided_backprop_saliency.GetSmoothedMask(image, feed_dict={...})

# Compute a 2D tensor for visualization.
grayscale_visualization = visualization.VisualizeImageGrayscale(
    smoothgrad_guided_backprop)

This is not an official Google product.