Our work Improving Semantic Image Segmentation via Label Fusion in Semantically Textured Meshes has been accepted at VISAPP 2022.