[Retracted] Hybrid Depth-Separable Residual Networks for Hyperspectral Image Classification

Table 3

Accuracy of different spatial sizes.

Spatial size	IN datasets			UP datasets			SA datasets
	OA (%)	AA (%)	Kappa (%)	OA (%)	AA (%)	Kappa (%)	OA (%)	AA (%)	Kappa (%)
5 × 5	95.86	96.08	95.09	98.36	98.20	98.25	98.89	98.88	98.85
7 × 7	98.96	98.25	98.64	99.29	99.35	99.27	99.58	99.45	99.57
9 × 9	99.30	99.31	99.24	99.68	99.66	99.58	99.87	99.86	99.88
11 × 11	99.72	99.60	99.70	99.86	99.83	99.81	99.98	99.98	99.98
13 × 13	99.65	99.61	99.60	99.87	99.83	99.85	100	100	100