Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Generating Satellite Imagery Data for Wildfire Detection through Mask-Conditioned Generative AI

Valeria Martin, K. Brent Venable, Derek Morgan

Abstract

The scarcity of labeled satellite imagery remains a fundamental bottleneck for deep-learning (DL)-based wildfire monitoring systems. This paper investigates whether a diffusion-based foundation model for Earth Observation (EO), EarthSynth, can synthesize realistic post-wildfire Sentinel-2 RGB imagery conditioned on existing burn masks, without task-specific retraining. Using burn masks derived from the CalFireSeg-50 dataset (Martin et al., 2025), we design and evaluate six controlled experimental configurations that systematically vary: (i) pipeline architecture (mask-only full generation vs. inpainting with pre-fire context), (ii) prompt engineering strategy (three hand-crafted prompts and a VLM-generated prompt via Qwen2-VL), and (iii) a region-wise color-matching post-processing step. Quantitative assessment on 10 stratified test samples uses four complementary metrics: Burn IoU, burn-region color distance (ΔC_burn), Darkness Contrast, and Spectral Plausibility. Results show that inpainting-based pipelines consistently outperform full-tile generation across all metrics, with the structured inpainting prompt achieving the best spatial alignment (Burn IoU = 0.456) and burn saliency (Darkness Contrast = 20.44), while color matching produces the lowest color distance (ΔC_burn = 63.22) at the cost of reduced burn saliency. VLM-assisted inpainting is competitive with hand-crafted prompts. These findings provide a foundation for incorporating generative data augmentation into wildfire detection pipelines. Code and experiments are available at: https://www.kaggle.com/code/valeriamartinh/genai-all-runned

Generating Satellite Imagery Data for Wildfire Detection through Mask-Conditioned Generative AI

Abstract

The scarcity of labeled satellite imagery remains a fundamental bottleneck for deep-learning (DL)-based wildfire monitoring systems. This paper investigates whether a diffusion-based foundation model for Earth Observation (EO), EarthSynth, can synthesize realistic post-wildfire Sentinel-2 RGB imagery conditioned on existing burn masks, without task-specific retraining. Using burn masks derived from the CalFireSeg-50 dataset (Martin et al., 2025), we design and evaluate six controlled experimental configurations that systematically vary: (i) pipeline architecture (mask-only full generation vs. inpainting with pre-fire context), (ii) prompt engineering strategy (three hand-crafted prompts and a VLM-generated prompt via Qwen2-VL), and (iii) a region-wise color-matching post-processing step. Quantitative assessment on 10 stratified test samples uses four complementary metrics: Burn IoU, burn-region color distance (ΔC_burn), Darkness Contrast, and Spectral Plausibility. Results show that inpainting-based pipelines consistently outperform full-tile generation across all metrics, with the structured inpainting prompt achieving the best spatial alignment (Burn IoU = 0.456) and burn saliency (Darkness Contrast = 20.44), while color matching produces the lowest color distance (ΔC_burn = 63.22) at the cost of reduced burn saliency. VLM-assisted inpainting is competitive with hand-crafted prompts. These findings provide a foundation for incorporating generative data augmentation into wildfire detection pipelines. Code and experiments are available at: https://www.kaggle.com/code/valeriamartinh/genai-all-runned

Paper Structure

This paper contains 35 sections, 14 equations, 7 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Background
  3. Denoising Diffusion Probabilistic Models
  4. Forward Process
  5. Reverse Process and Training
  6. Conditioning
  7. Inpainting for Controlled Synthesis
  8. Generating Images from Burn Masks (EarthSynth)
  9. Vision Language Models for Prompt Generation
  10. Related Work
  11. Methods
  12. Experimental Dataset
  13. EarthSynth Generative Framework
  14. Pipeline Configurations
  15. Prompt Engineering Strategies
  16. ...and 20 more sections

Figures (7)

  • Figure 1: Metric distributions per experiment (all prompt strategies pooled). Boxes show median $\pm$ IQR across 10 test samples.
  • Figure 2: Mean metrics by experiment $\times$ prompt strategy.
  • Figure 3: Visual Results for Sample S00 (burn ratio 10%)
  • Figure 4: Visual Results for Sample S02 (burn ratio 30%)
  • Figure 5: Visual Results for Sample S05 (burn ratio 50%)
  • ...and 2 more figures