Search papers, labs, and topics across Lattice.
The paper introduces MorphoFormer, a framework for joint building height (BH) and building footprint (BF) estimation from remote sensing data, which explicitly encodes cross-task coupling via a BF-Guided Task Decoder (BGTD) and a Morphology Consistency Loss (MCL). BGTD gates the height branch using cross-attention on footprint-derived morphology, while MCL supervises a height-from-footprint surrogate to enforce BF feature encoding of height-correlated structure. Experiments on a dataset of 54 cities demonstrate that MorphoFormer reduces BH test RMSE by 0.24m compared to a Swin-MTL baseline, with ablations showing that both BGTD and MCL contribute significantly to the improvement.
Encoding cross-task relationships between building footprints and heights slashes height estimation error by 7% – more effective than just refining individual encoders.
Building height (BH) and building footprint (BF) jointly describe the vertical and horizontal extent of the built environment and are required inputs for urban climate, disaster-risk, and population-mapping models. The two parameters are coupled through floor-area-ratio (FAR) constraints, yet remote-sensing approaches typically treat them as independent regression targets. We argue that explicitly encoding this cross-task coupling is more impactful than further refining individual encoders, and propose MorphoFormer, a joint BH/BF estimation framework built around two complementary mechanisms: (i) a BF-Guided Task Decoder (BGTD) that gates the height branch via cross-attention on a footprint-derived morphology context, and (ii) a Morphology Consistency Loss (MCL) that supervises a height-from-footprint surrogate against the ground-truth BH, indirectly forcing the BF feature to encode height-correlated structure. The encoder is a single-stage Swin backbone fed by Sentinel-1 SAR, Sentinel-2 multispectral, and DEM inputs, trained and evaluated on a geo-blocked split of 54 cities. Against a Swin-MTL baseline at identical receptive field, MorphoFormer reduces BH test RMSE from 3.39 to 3.15 m (R^2 improves 0.62 ->0.67) with BF R^2 stable at 0.80. Controlled ablations at identical capacity attribute most of this 0.24 m improvement to the two proposed mechanisms: removing BGTD raises BH RMSE by 0.11 m and removing MCL raises it by 0.11 m, with the residual approximately 0.02 m falling within the noise floor of encoder-side variations. Because both mechanisms act on cross-task representations rather than pixels, the design carries no intrinsic dependence on input resolution.
