Fig. 9
From: Semantic segmentation and photogrammetry of crowdsourced images to monitor historic facades
![Fig. 9](http://media.springernature.com/full/springer-static/image/art%3A10.1186%2Fs40494-022-00664-y/MediaObjects/40494_2022_664_Fig9_HTML.png)
The boxplot of the results of one-stage model and two-stage model’s performance on overall classes and plant class, where red triangles represent mean values. IoU and F1-score are not accounted if there are no groundtruth labels for that class in the image