A Large Scale Open-Source Image and Video Dataset for Robust Wildfire Detection and Classification

2026-06-08

AI Summary

This study addresses the limitations of existing wildfire detection models, which are hindered by the absence of large-scale, geographically diverse open-source datasets encompassing complex environmental conditions. To bridge this gap, we introduce the Global Wildfire Prevention Dataset (GWFP), the first to integrate multispectral imagery, real-world negative samples, and meteorological interferences such as mist and haze, covering flame, smoke, ember, and near-infrared modalities. Leveraging this dataset, we propose a lightweight HTE-ResNet architecture that enables efficient interaction between frequency- and spatial-domain features through Hadamard-enhanced residual connections. Experimental results demonstrate that our approach significantly improves model generalization across domains, underscoring GWFP’s critical role in enhancing the robustness and practicality of wildfire monitoring systems.

Abstract

Wildfire detection and monitoring are critical for mitigating fire spread and reducing environmental and infrastructural damage. In this work, we introduce GWFP (Global Wildfire Prevention Dataset), a large-scale, open-source dataset of wildfire images and videos designed to support early fire and smoke detection research. GWFP contains geographically diverse wildfire scenes, including flames, smoke, Waterdog/Fog environmental conditions, Near Infrared (NIR) imagery, Ember, and challenging negative samples collected from real-world scenarios worldwide. To evaluate dataset robustness and cross-domain generalization, we benchmark multiple convolutional and transformer-based architectures across both in-domain and cross-dataset settings. Additionally, we explore lightweight frequency--spatial feature interaction using Hadamard-enhanced residual connections (HTE-ResNet) to analyze representation robustness under domain-shift conditions. Experimental results demonstrate strong cross-dataset generalization and practical utility for real-world wildfire monitoring applications. The dataset and source code will be publicly released upon acceptance.

Problem

• wildfire detection
• smoke detection
• open-source dataset
• cross-domain generalization
• robust classification

Innovation

• wildfire detection
• large-scale dataset
• cross-domain generalization
• frequency-spatial feature interaction
• Hadamard-enhanced residual connections