Weakly-supervised fire segmentation by visualizing intermediate CNN layers

TL;DR

This paper introduces a weakly-supervised fire segmentation method using CNN intermediate layer visualization, achieving better accuracy than traditional CAM by leveraging mean feature values and rotation equivariant regularization.

Contribution

It proposes a novel weakly-supervised approach for fire segmentation that outperforms conventional CAM methods by utilizing mean features and a rotation regularization loss.

Findings

Mean feature values outperform CAM in fire segmentation

Rotation equivariant regularization improves segmentation accuracy

Noticeable improvements over baseline methods

Abstract

Fire localization in images and videos is an important step for an autonomous system to combat fire incidents. State-of-art image segmentation methods based on deep neural networks require a large number of pixel-annotated samples to train Convolutional Neural Networks (CNNs) in a fully-supervised manner. In this paper, we consider weakly supervised segmentation of fire in images, in which only image labels are used to train the network. We show that in the case of fire segmentation, which is a binary segmentation problem, the mean value of features in a mid-layer of classification CNN can perform better than conventional Class Activation Mapping (CAM) method. We also propose to further improve the segmentation accuracy by adding a rotation equivariant regularization loss on the features of the last convolutional layer. Our results show noticeable improvements over baseline method for weakly-supervised fire segmentation.