Seeing Heat with Color -- RGB-Only Wildfire Temperature Inference from SAM-Guided Multimodal Distillation using Radiometric Ground Truth

TL;DR

This paper presents SAM-TIFF, a novel RGB-only wildfire temperature prediction framework that distills knowledge from multimodal data, enabling cost-effective UAV wildfire monitoring without thermal sensors.

Contribution

It introduces a teacher-student distillation approach for pixel-level temperature inference using RGB data, eliminating the need for thermal sensors in wildfire monitoring.

Findings

Achieves accurate per-pixel temperature regression from RGB UAV data.

Demonstrates strong generalization on the FLAME 3 dataset.

Enables lightweight wildfire monitoring systems without thermal sensors.

Abstract

High-fidelity wildfire monitoring using Unmanned Aerial Vehicles (UAVs) typically requires multimodal sensing - especially RGB and thermal imagery - which increases hardware cost and power consumption. This paper introduces SAM-TIFF, a novel teacher-student distillation framework for pixel-level wildfire temperature prediction and segmentation using RGB input only. A multimodal teacher network trained on paired RGB-Thermal imagery and radiometric TIFF ground truth distills knowledge to a unimodal RGB student network, enabling thermal-sensor-free inference. Segmentation supervision is generated using a hybrid approach of segment anything (SAM)-guided mask generation, and selection via TOPSIS, along with Canny edge detection and Otsu's thresholding pipeline for automatic point prompt selection. Our method is the first to perform per-pixel temperature regression from RGB UAV data, demonstrating strong generalization on the recent FLAME 3 dataset. This work lays the foundation for lightweight, cost-effective UAV-based wildfire monitoring systems without thermal sensors.