Fig. 8From: Semantic segmentation and photogrammetry of crowdsourced images to monitor historic facadesThe visual representation of the segmented photographs. The first row are the results of the one-stage model, and the second row are the results of two-stage modelBack to article page