Calligraphy and Painting Identification 3D-CNN Model Based on Hyperspectral Image MNF Dimensionality Reduction

Table 1

Results of calligraphy and painting author identification of different CNN models with different kinds of input.

	Training set
Network model	3D-CNN based on multichannel dimensionality reduced images	2D-CNN based on MNF pseudocolor image mosaic as input	2D-CNN based on multichannel dimensionality reduced images directly as input
Accuracy	1.000	1.000	1.000
	Test set
Network model	3D-CNN based on multichannel dimensionality reduced images	2D-CNN based on MNF pseudocolor image mosaic as input	2D-CNN based on multichannel dimensionality reduced images directly as input
Accuracy	0.932	0.926	0.937