SupplementalMaterialsfor" HardExample GenerationbyTextureSynthesisforCross-domain ShapeSimilarityLearning "

The network architectures for our geometry-focused multi-view metric learning framework are reportedinTable1(TSM),Table2(AMV-ML),andFigure1. In Table 3, we report the statistics of the used datasets in our submission. Note that, 3D-FUTURE here contains more fine-grained 3DCAD models.