Detecting Wildfires on UAVs with Real-time Segmentation Trained by Larger Teacher Models

TL;DR

This paper presents a real-time wildfire smoke segmentation method for UAVs that uses small models trained with bounding box labels and zero-shot supervision, enabling effective onboard detection during forest fires.

Contribution

It introduces a novel training approach for small wildfire smoke segmentation models using only bounding box labels and foundation model supervision.

Findings

Achieved 63.3% mIoU on a wildfire dataset.

Runs at ~25 fps on NVIDIA Jetson Orin NX.

Successfully demonstrated real-world wildfire detection.

Abstract

Early detection of wildfires is essential to prevent large-scale fires resulting in extensive environmental, structural, and societal damage. Uncrewed aerial vehicles (UAVs) can cover large remote areas effectively with quick deployment requiring minimal infrastructure and equipping them with small cameras and computers enables autonomous real-time detection. In remote areas, however, detection methods are limited to onboard computation due to the lack of high-bandwidth mobile networks. For accurate camera-based localisation, segmentation of the detected smoke is essential but training data for deep learning-based wildfire smoke segmentation is limited. This study shows how small specialised segmentation models can be trained using only bounding box labels, leveraging zero-shot foundation model supervision. The method offers the advantages of needing only fairly easily obtainable bounding box labels and requiring training solely for the smaller student network. The proposed method achieved 63.3% mIoU on a manually annotated and diverse wildfire dataset. The used model can perform in real-time at ~25 fps with a UAV-carried NVIDIA Jetson Orin NX computer while reliably recognising smoke, as demonstrated at real-world forest burning events. Code is available at: https://gitlab.com/fgi_nls/public/wildfire-real-time-segmentation