Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FLAME Diffuser: Wildfire Image Synthesis using Mask Guided Diffusion

Hao Wang, Sayed Pedram Haeri Boroujeni, Xiwen Chen, Ashish Bastola, Huayu Li, Wenhui Zhu, Abolfazl Razi

TL;DR

The FLAME Diffuser is presented, a training-free, diffusion-based framework designed to generate realistic wildfire images with paired ground truth, and the fusion of Perlin noise in this work significantly improved the quality of synthesized images.

Abstract

Wildfires are a significant threat to ecosystems and human infrastructure, leading to widespread destruction and environmental degradation. Recent advancements in deep learning and generative models have enabled new methods for wildfire detection and monitoring. However, the scarcity of annotated wildfire images limits the development of robust models for these tasks. In this work, we present the FLAME Diffuser, a training-free, diffusion-based framework designed to generate realistic wildfire images with paired ground truth. Our framework uses augmented masks, sampled from real wildfire data, and applies Perlin noise to guide the generation of realistic flames. By controlling the placement of these elements within the image, we ensure precise integration while maintaining the original images style. We evaluate the generated images using normalized Frechet Inception Distance, CLIP Score, and a custom CLIP Confidence metric, demonstrating the high quality and realism of the synthesized wildfire images. Specifically, the fusion of Perlin noise in this work significantly improved the quality of synthesized images. The proposed method is particularly valuable for enhancing datasets used in downstream tasks such as wildfire detection and monitoring.

FLAME Diffuser: Wildfire Image Synthesis using Mask Guided Diffusion

TL;DR

The FLAME Diffuser is presented, a training-free, diffusion-based framework designed to generate realistic wildfire images with paired ground truth, and the fusion of Perlin noise in this work significantly improved the quality of synthesized images.

Abstract

Wildfires are a significant threat to ecosystems and human infrastructure, leading to widespread destruction and environmental degradation. Recent advancements in deep learning and generative models have enabled new methods for wildfire detection and monitoring. However, the scarcity of annotated wildfire images limits the development of robust models for these tasks. In this work, we present the FLAME Diffuser, a training-free, diffusion-based framework designed to generate realistic wildfire images with paired ground truth. Our framework uses augmented masks, sampled from real wildfire data, and applies Perlin noise to guide the generation of realistic flames. By controlling the placement of these elements within the image, we ensure precise integration while maintaining the original images style. We evaluate the generated images using normalized Frechet Inception Distance, CLIP Score, and a custom CLIP Confidence metric, demonstrating the high quality and realism of the synthesized wildfire images. Specifically, the fusion of Perlin noise in this work significantly improved the quality of synthesized images. The proposed method is particularly valuable for enhancing datasets used in downstream tasks such as wildfire detection and monitoring.
Paper Structure (19 sections, 3 equations, 9 figures, 2 tables)

This paper contains 19 sections, 3 equations, 9 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Wildfire Datasets
  4. Latent Diffusion
  5. Image-Text Architecture
  6. Problem Formulation
  7. Limitations of Existing Image-to-Image Diffusion
  8. Image Manipulation
  9. Methodology
  10. Mask Generation
  11. Mask-Image Diffusion
  12. Experiments
  13. Qualitative Evaluation
  14. Quantitative Evaluation
  15. Ablation Study
  16. ...and 4 more sections

Figures (9)

  • Figure 1: The concept of wildfire image synthesis, where the masks can guide the location of flames in the generated images.
  • Figure 2: Sample images of the FLAME1, FLAME2, and DFire datasets.
  • Figure 3: Framework of the FLAME Diffuser.
  • Figure 4: Wildfire image synthesis in terms of text prompt scale and image denoise strength. Text prompt: 'Wildfires in the snow, flame in the smoke, high-resolution, photo realistic.'
  • Figure 5: Sample images from FLAME diffuser. The top row is the input mask, and left leftmost column is the input style image. (a) and (b) are two different scenes from the FLAME2 dataset, (c) and (d) are another two scenes from the FLAME1 dataset.
  • ...and 4 more figures