Reconstructing 3D Model from Single-View Sketch with Deep Neural Network

Table 3

Character models (synthetic).

	Ours	ShapeMVD	Nearest retrieval	Tatarchenko et al. [15]	[15]+U-net	Volumetric decoder	R2N2
Hausdorff distance	0.065	0.089	0.200	0.119	0.092	0.152	0.148
Chamfer distance	0.010	0.015	0.036	0.025	0.016	0.026	0.032
Normal distance	26.47	30.61	44.93	34.98	31.00	53.84	53.13
Depth map error	0.014	0.018	0.040	0.030	0.019	0.031	0.036
Volumetric distance	0.248	0.313	0.541	0.428	0.329	0.437	0.493