Image Classification on CIFAR-100-C mixture of 15 corruptions (val) benchmark leaderboard · SOTA2 Research

Image Classification on CIFAR-100-C mixture of 15 corruptions (val)

30.3Gaussian Noise AccuracyRamen

Evaluation Results

Method	Links
Ramen2026.04		30.3	32.9	42.8	52.5	29.1	52.2	55.5	55.4	55.1	52.6	65.9	51	36.8	40.4	39.6	46.1
SAR2026.04		28.4	30.5	38.5	50.4	24.6	49.2	55.1	54.1	53.4	47.2	64	45	33.6	29.7	37.4	42.7
WATT-S2026.04		26.4	28.4	35.6	50.2	24.2	48.5	54.4	54.2	54	47.7	63.5	43.3	34	31.4	37.7	42.2
NOTE2026.04		25.7	27.3	35.4	49.9	23.9	47.6	54.2	52.7	51.8	45.9	63.7	44	32.4	30.2	36	41.4
Tent2026.04		24.9	26.7	34.4	49.7	23.9	47.5	53.9	52.9	51.4	45.3	62.9	43.6	32	31.7	37.1	41.2
TDA2026.04		23.4	25.2	30.3	44.1	20.3	43.8	49.1	49.3	51.3	42.2	58.7	36.1	29.3	26.4	33.6	37.5
Ensemble2026.04		22.9	24.4	29.6	43.6	20.1	43.7	48.7	48.9	50.4	41.8	58.1	35.3	29.2	26.3	33.6	37.1
RoTTA2026.04		22.9	24.5	32.4	47.9	25.4	46.5	50.1	50.5	50.7	50.8	58.3	53.1	37.7	29.9	36.2	41.1
DMN-ZS2026.04		22.9	24.4	29.6	44.3	20.2	44.1	49.4	49.7	51.1	42.2	58.8	35.5	29.4	26.2	34	37.5
Mint2026.04		22.3	24.5	33	49.2	22.4	47.8	54.2	52.4	51.6	47.7	63.7	42.2	31.8	24	34.3	40.1
CLIPArTT2026.04		21.1	22.8	29.4	46.3	24.5	45	51.1	51.5	51.8	44	61.1	39	33.2	28.3	36.7	39
CLIP2026.04		19.7	21.4	25.3	42.5	20.2	43.1	48	48.4	49.7	41.7	57	34.5	29.2	23.9	32.4	35.8