GeoDiT: Point-Conditioned Diffusion Transformer for Satellite Image Synthesis
This addresses the need for more efficient and semantically rich control in satellite image generation for applications like remote sensing, though it is incremental as it builds on existing diffusion and transformer methods.
The paper tackles the problem of generating satellite images with flexible control by introducing GeoDiT, a diffusion transformer that uses point-based conditioning instead of pixel-level maps, achieving state-of-the-art performance in remote sensing image synthesis.
We introduce GeoDiT, a diffusion transformer designed for text-to-satellite image generation with point-based control. Existing controlled satellite image generative models often require pixel-level maps that are time-consuming to acquire, yet semantically limited. To address this limitation, we introduce a novel point-based conditioning framework that controls the generation process through the spatial location of the points and the textual description associated with each point, providing semantically rich control signals. This approach enables flexible, annotation-friendly, and computationally simple inference for satellite image generation. To this end, we introduce an adaptive local attention mechanism that effectively regularizes the attention scores based on the input point queries. We systematically evaluate various domain-specific design choices for training GeoDiT, including the selection of satellite image representation for alignment and geolocation representation for conditioning. Our experiments demonstrate that GeoDiT achieves impressive generation performance, surpassing the state-of-the-art remote sensing generative models.