Search papers, labs, and topics across Lattice.
This paper introduces FlowForm, a novel framework for synthesizing satellite flood imagery by integrating Shallow Water Equation-inspired latent regularization with structure-aware conditioning. By leveraging a Flood Descriptor Module (FDM) and a Terrain Anchor Adapter (TAA), FlowForm enhances the spatial accuracy and visual fidelity of generated flood images, addressing the limitations of existing generative models. The authors validate their approach using a newly curated dataset, FloodScape, demonstrating superior performance in visual quality and consistency of flooded regions compared to traditional methods.
FlowForm achieves unprecedented visual fidelity and consistency in satellite flood imagery synthesis, outperforming existing generative models.
Developing robust flood assessment models requires high-quality paired satellite imagery, yet such data remain scarce for flood-specific image generation. Although generative models provide a promising means of data augmentation, existing methods often yield implausible spatial layouts of flooded regions and distort scene structures. We propose FlowForm, a framework for satellite flood synthesis that integrates SWE-inspired latent regularization with structure-aware conditioning. The Flood Descriptor Module (FDM) imposes differentiable penalties on residuals of the steady-state Shallow Water Equation in auxiliary latent fields at the diffusion bottleneck. The Terrain Anchor Adapter (TAA) injects depth, semantic, and edge features at four encoder scales of the U-Net. We further curate FloodScape, a large-scale, high-resolution dataset comprising paired satellite images acquired before and after disasters. In addition to standard image-generation metrics, we evaluate the consistency of flooded regions, zero-shot generalization to a geographically held-out flood event, and sensitivity to individual components. Across all reported comparisons, FlowForm achieves higher visual fidelity, greater similarity between paired images, and stronger consistency of flooded regions.
