Attention on Classification for Fire Segmentation

TL;DR

This paper introduces a CNN with spatial and channel attention mechanisms for joint fire classification and segmentation, enhancing fire detection accuracy by capturing global context and leveraging classification probabilities.

Contribution

It proposes a novel CNN architecture that combines spatial self-attention and a classification-based channel attention module for improved fire segmentation.

Findings

Enhanced segmentation accuracy over existing methods.

Effective use of classification probability as attention weight.

Joint training improves both classification and segmentation performance.

Abstract

Detection and localization of fire in images and videos are important in tackling fire incidents. Although semantic segmentation methods can be used to indicate the location of pixels with fire in the images, their predictions are localized, and they often fail to consider global information of the existence of fire in the image which is implicit in the image labels. We propose a Convolutional Neural Network (CNN) for joint classification and segmentation of fire in images which improves the performance of the fire segmentation. We use a spatial self-attention mechanism to capture long-range dependency between pixels, and a new channel attention module which uses the classification probability as an attention weight. The network is jointly trained for both segmentation and classification, leading to improvement in the performance of the single-task image segmentation methods, and the previous methods proposed for fire segmentation.