Search papers, labs, and topics across Lattice.
This paper introduces Genesis, a generative engine designed for multi-scale tile completion of satellite imagery, addressing the limitations of existing models that either enhance resolution or extend imagery at a fixed scale. By combining vertical super-resolution and horizontal mask-based outpainting, Genesis synthesizes complete and consistent quadtree pyramids from sparse seed tiles across varying zoom levels and spatial contexts. The results demonstrate state-of-the-art performance in generating seamless multi-resolution maps, validated using the newly introduced dense500 dataset, which encompasses diverse geographic regions.
Genesis achieves seamless multi-resolution satellite imagery synthesis, unifying super-resolution and outpainting for globally consistent quadtree generation.
Earth observation is fundamentally multi-scale; geospatial tasks span varied resolutions, and satellite imagery is organized into cascading tile pyramids that nest fine detail within wide coverage. Current generative models of satellite imagery, however, operate along a single axis: they either zoom to enhance a single tile's resolution or pan to extend imagery at a fixed scale. As a result, no existing method produces a complete pyramid that stays consistent across both scale and space, where a high-zoom tile must agree with the coarse context it refines and with the neighbors it meets. Motivated by this gap, we introduce a new task, multi-scale tile completion: given a sparse set of seed tiles at arbitrary zoom levels and positions, synthesize a complete, uniform quadtree that is globally consistent across both scale and space. We approach this task with Genesis, a generative engine that brings both axes together by composing two specialized operators over the quadtree: a vertical super-resolution model and a horizontal mask-based outpainting model, producing pyramids that are consistent across zoom levels and seamless across neighboring tiles. Each operator achieves state-of-the-art results on its subtask, and the engine propagates sparse seeds into seamless, multi-resolution maps from any initial configuration. To evaluate the task and benchmark Genesis, we introduce dense500, a fully observed multi-scale pyramid dataset spanning diverse geographic regions, together with a suite of pyramid-level metrics. Code, models, and our dataset are available at https://github.com/mvrl/genesis.
