Abs-CAM: A Gradient Optimization Interpretable Approach for Explanation of Convolutional Neural Networks

by   Chunyan Zeng, et al.

The black-box nature of Deep Neural Networks (DNNs) severely hinders its performance improvement and application in specific scenes. In recent years, class activation mapping-based method has been widely used to interpret the internal decisions of models in computer vision tasks. However, when this method uses backpropagation to obtain gradients, it will cause noise in the saliency map, and even locate features that are irrelevant to decisions. In this paper, we propose an Absolute value Class Activation Mapping-based (Abs-CAM) method, which optimizes the gradients derived from the backpropagation and turns all of them into positive gradients to enhance the visual features of output neurons' activation, and improve the localization ability of the saliency map. The framework of Abs-CAM is divided into two phases: generating initial saliency map and generating final saliency map. The first phase improves the localization ability of the saliency map by optimizing the gradient, and the second phase linearly combines the initial saliency map with the original image to enhance the semantic information of the saliency map. We conduct qualitative and quantitative evaluation of the proposed method, including Deletion, Insertion, and Pointing Game. The experimental results show that the Abs-CAM can obviously eliminate the noise in the saliency map, and can better locate the features related to decisions, and is superior to the previous methods in recognition and localization tasks.


page 5

page 6

page 7


Enhancing Deep Neural Network Saliency Visualizations with Gradual Extrapolation

We propose an enhancement technique of the Class Activation Mapping meth...

Shap-CAM: Visual Explanations for Convolutional Neural Networks based on Shapley Value

Explaining deep convolutional neural networks has been recently drawing ...

Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks

Recently, increasing attention has been drawn to the internal mechani...

IDGI: A Framework to Eliminate Explanation Noise from Integrated Gradients

Integrated Gradients (IG) as well as its variants are well-known techniq...

Improving Deep Learning Interpretability by Saliency Guided Training

Saliency methods have been widely used to highlight important input feat...

Verifying Attention Robustness of Deep Neural Networks against Semantic Perturbations

It is known that deep neural networks (DNNs) classify an input image by ...

VS-CAM: Vertex Semantic Class Activation Mapping to Interpret Vision Graph Neural Network

Graph convolutional neural network (GCN) has drawn increasing attention ...

Please sign up or login with your details

Forgot password? Click here to reset