Hyperspectral Image Classification on Pavia University (5% samples per class) benchmark leaderboard · SOTA2 Research

Hyperspectral Image Classification on Pavia University (5% samples per class)

99.92Overall AccuracySCT-Net

Evaluation Results

Method	Links
SCT-Net2026.04		99.92	99.88	99.89	99.87	100	99.85	99.69	100	100	100	99.69	99.89
CNN-MRF2017.05		99.71	99.55	99.63	—	—	—	—	—	—	—	—	—
SSFTT2026.04		99.7	99.4	99.61	99.94	99.97	98.97	98.81	99.8	99.96	99.92	99.14	98.05
DC-CNN2017.05		99.68	99.5	99.58	—	—	—	—	—	—	—	—	—
MASSFormer2026.04		99.48	99.06	99.31	99.37	99.99	96.91	98.76	100	99.85	99.92	98.52	98.66
HybridSN2026.04		98.16	97.35	97.57	96.51	94.49	94.18	99.55	96.71	99.43	100	95.97	95.55
SS-Mamba2026.04		95.78	98.31	94.51	95.81	92.54	99.7	99.14	100	99	99.98	99.72	99.99
SPP-DCNN2017.05		94.88	93.29	93.21	—	—	—	—	—	—	—	—	—
SS-DCNN2017.05		93.34	92.2	91.95	—	—	—	—	—	—	—	—	—
SpectralFormer2026.04		93.17	90.27	91.23	96.9	95.59	78.9	86.31	93.91	82.54	88.9	95.17	93.97
3D-CNN2026.04		92.01	90.45	90.87	93.3	94.21	90.19	91.29	95.47	93.86	81.84	80.36	92.58
2D-CNN2026.04		90.19	87.31	87.52	93.28	94.9	75.55	93.87	97.88	70.05	70.92	90.96	90.78
SVM2026.04		82.76	71.44	76.28	91.22	97.78	34.95	81.55	98.69	41.5	32.17	89.74	96.87