Info Results – DGCNN-Mod – A

Further info

Authors	(preliminary tests) R. Pierdicca, M. Paolanti, F. Matrone, M. Martini, C. Morbidoni, E. S. Malinverni, E. Frontoni, A. Lingua (with ArCH dataset) F. Matrone, E. Grilli, M. Martini, M. Paolanti, R. Pierdicca, F. Remondino
Institution	Università Politecnica delle Marche, Politecnico di Torino, FBK
Approach	Deep learning, neural network
Name	DGCNN-Mod
Short description	The K-nn input features have been modified with the addition of the r, g, b and normals
Title of the paper	(preliminary tests) Point Cloud Semantic Segmentation Using a Deep Learning Framework for Cultural Heritage (with ArCH dataset) Comparing Machine and Deep Learning Methods for Large 3D Heritage Semantic Segmentation
Reference link	https://www.mdpi.com/2072-4292/12/6/1005
Code	–
Additional features	–
GPU/Hardware	Nvidia RTX 2080 TI 11 GB, 128 GB RAM, processor Intel(R) Xeon(R) Silver 4214 CPU @ 2.20GHz
Linked projects	Partially funded by the CIVITAS (ChaIn for excellence of reflectiVe societies to exploit dIgital culTural heritAge and museumS) project and by external funding from the project “Artificial Intelligence for Cultural Heritage” (AI4CH) joint Italy-Israel lab which was funded by the Italian Ministry of Foreign Affairs and International Cooperation (MAECI)
Comments	–
Date of publication in this website	September 2020

	Arch	Column	Molding	Floor	Door-Window	Wall	Stair	Vault	Roof
Precision	0.289	0.763	0.414	0.825	0.779	0.787	0.801	0.792	0.950
Recall	0.080	0.191	0.360	0.874	0.178	0.927	0.498	0.951	0.936
F1-Score	0.126	0.307	0.385	0.849	0.289	0.851	0.616	0.864	0.942
IoU	0.067	0.180	0.238	0.737	0.169	0.741	0.443	0.761	0.892