Image Classification on CIFAR-100 (test) (Symmetric/Asymmetric/Instance-dependent Accuracy η=0.2, 0.4, 0.6) benchmark leaderboard · SOTA2 Research

Image Classification on CIFAR-100 (test) (Symmetric/Asymmetric/Instance-dependent Accuracy η=0.2, 0.4, 0.6)

68.73Test Accuracy (Symmetric, η=0.2)Robust Self-Training with Closed-loop Label Correction

Evaluation Results

Method
Robust Self-Training with Closed-loop Label Correction2026.03	68.73	63	50.94	70.18	67.21	63.8	70.69	63.72	55.08
EMLC2026.03	68.67	62.49	46.9	70.1	65.06	57.15	63.82	53.99	32.11
Robust Self-Training with Closed-loop Label Correction2026.03	68.47	62.36	50.74	69.76	66.76	63.22	70.64	63.58	54.97
FasTEN2026.03	68.09	62.27	50.42	67.64	53.42	31.5	66.57	59.29	47.39
MW-Net2026.03	67.94	62.26	54.24	70.04	65.26	57.23	62.87	48.31	26.53
Cross-Entropy2026.03	67.59	59.99	49.94	69.69	65.33	55.96	62.57	45.9	24.05
L2B2026.03	64.57	54.54	20.66	64.91	58.42	43.33	66.27	59.55	47.45
L2B2026.03	64.28	54.38	20.66	64.66	58.11	43.33	66.05	59.55	47.45
Cross-Entropy2026.03	62.93	45.8	26.81	62.8	46.62	30.75	60.01	40.09	20.88
EMLC2026.03	62.8	46.51	31.74	62.94	47.09	32.34	61.94	49.34	28.44
MW-Net2026.03	62.38	48.59	29.55	62.87	47.82	31.95	59.25	40.91	22.75
FasTEN2026.03	59.37	43.65	26.09	61.91	43.27	31.5	59.7	45.79	33.15