Image Classification on CIFAR-100 (Provable Robustness Evaluation) benchmark leaderboard · SOTA2 Research

Image Classification on CIFAR-100 (Provable Robustness Evaluation)

51.84Standard AccuracyCRC-Lip

Evaluation Results

Method
CRC-Lip2022.11	51.84	38.54	27.32	18.53	4.47	4.55
CRC-Lip2022.11	51.59	39.27	27.94	19.06	5.82	7.19
CRC-Lip2022.11	51.42	39.01	28.94	20.29	8	8.65
CRC-Lip2022.11	50.97	38.77	27.73	19.28	4.58	5.69
CRC-Lip2022.11	50.79	37.5	26.16	17.27	2.73	2.98
CRC-Lip2022.11	50.11	38.69	28.45	20.05	8.39	10.16
Baseline2022.11	48.06	34.52	23.08	14.7	—	—
SOC faster gradient computation2022.11	47.97	33.84	22.66	14.26	-0.09	-0.68
Baseline2022.11	47.37	33.99	23.4	14.69	—	—
SOC faster gradient computation2022.11	46.41	33.07	22.06	14	-0.96	-0.92
Baseline2022.11	46.39	33.08	22.02	13.77	—	—
SOC faster gradient computation2022.11	45.86	32.54	21.18	12.77	-0.53	-0.54
Baseline2022.11	45.77	32.08	21.36	13.64	—	—
SOC faster gradient computation2022.11	45.28	31.67	20.69	13.26	-0.49	-0.41
Baseline2022.11	43.42	30.36	19.71	12.66	—	—
SOC faster gradient computation2022.11	42.78	29.88	19.73	12.52	-0.64	-0.48
SOC faster gradient computation2022.11	42.07	28.51	18.86	11.89	0.35	-0.02
Baseline2022.11	41.72	28.53	18.37	11.49	—	—