U-Net with Hadamard Transform and DCT Latent Spaces for Next-day Wildfire Spread Prediction

2026-02-12

AI Summary

This work addresses the challenge of balancing prediction accuracy and computational efficiency in next-day wildfire spread forecasting by proposing TD-FusionUNet, a lightweight model that achieves high performance with minimal parameters. The model introduces, for the first time, learnable Hadamard and discrete cosine transforms within the U-Net latent space to construct an orthogonal frequency-domain feature fusion mechanism. Combined with Gaussian mixture modeling and stochastic edge-cropping preprocessing, this approach significantly enhances representation capability for sparse pre-fire masks. Evaluated on multimodal satellite data—including the WildfireSpreadTS dataset—TD-FusionUNet attains an F1 score of 0.591 with only 370K parameters, substantially outperforming a ResNet18-U-Net baseline while maintaining low computational overhead.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Multi-modal Vision
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We developed a lightweight and computationally efficient tool for next-day wildfire spread prediction using multimodal satellite data as input. The deep learning model, which we call Transform Domain Fusion UNet (TD-FusionUNet), incorporates trainable Hadamard Transform and Discrete Cosine Transform layers that apply two-dimensional transforms, enabling the network to capture essential"frequency"components in orthogonalized latent spaces. Additionally, we introduce custom preprocessing techniques, including random margin cropping and a Gaussian mixture model, to enrich the representation of the sparse pre-fire masks and enhance the model's generalization capability. The TD-FusionUNet is evaluated on two datasets which are the Next-Day Wildfire Spread dataset released by Google Research in 2023, and WildfireSpreadTS dataset. Our proposed TD-FusionUNet achieves an F1 score of 0.591 with 370k parameters, outperforming the UNet baseline using ResNet18 as the encoder reported in the WildfireSpreadTS dataset while using substantially fewer parameters. These results show that the proposed latent space fusion model balances accuracy and efficiency under a lightweight setting, making it suitable for real time wildfire prediction applications in resource limited environments.

Problem

• wildfire spread prediction
• next-day forecasting
• multimodal satellite data
• lightweight model
• real-time prediction

Innovation

• Hadamard Transform
• Discrete Cosine Transform
• Latent Space Fusion
• Lightweight Deep Learning
• Wildfire Spread Prediction