Facades Material Munich (FaMatMuc) - Semantic Segmentation and Classification dataset

This dataset was developed and published in conjunction with the paper "Decoding Building Facades: Automated Element and Material Recognition in Street-Level Images for Large-Scale Building Stock Assessments" (see Links), presented at the International Workshop on Intelligent Computing in Engineering in Glasgow, 2025.
To support the proposed approaches, a dedicated material dataset with intra-urban facades in the city of Munich was created with two types of annotations, semantic segmentation masks and classification labels.The semantic segmentation dataset comprises 100 annotated high-resolution images (5304 x 7952 pixels); the classification dataset includes 541 close-up patches (256 x 256 pixels) extracted from these 100 images. This dataset primarily features single residential and office buildings taken from a frontal perspective. Annotated classes are Plaster, Brick, Stone, Concrete, Glass, Timber, and Other.