Enhancing Photogrammetric Digital Surface Models with Pretrained Diffusion Models and Multimodal Conditioning
A diffusion model cut dense-urban DSM error nearly in half in the paper’s French-city tests.
The authors use a modified Stable Diffusion 3 setup to refine noisy satellite photogrammetry DSMs with help from Pléiades imagery. They prune the text stream and add patch-wise normalization so the model can train on LiDAR elevation data. Reported Dense Urban RMSE falls from 6.00 to 3.45 m in in-context cities, and from 4.16 to 2.77 m in the held-out city of Bordeaux. HF Daily Papers' note
The authors use a modified Stable Diffusion 3 setup to refine noisy satellite photogrammetry DSMs with help from Pléiades imagery. They prune the text stream and add patch-wise normalization so the model can train on LiDAR elevation data. Reported Dense Urban RMSE falls from 6.00 to 3.45 m in in-context cities, and from 4.16 to 2.77 m in the held-out city of Bordeaux. HF Daily Papers' note
score 4