Share your thoughts, 1 month free Claude Pro on usSee more

Certified Robustness on CIFAR-10 (test)

94.5Accuracy (Standard)

Diffusion Calibration

Updated 4d ago

Evaluation Results

Method
Diffusion Calibration 2023.10	94.5	-	61.9	-	-	-	-	-	-	-	-	-	90.3	53.5	-	-	-	-
Carlini et al. 2023.10	92.8	-	59.8	-	-	-	-	-	-	-	-	-	89.5	52.7	-	-	-	-
Gaussian 2023.10	92.7	-	55.3	-	-	-	-	-	-	-	-	-	87	47.9	-	-	-	-
Cascaded Randomized Smoothing 2023.10	92.5	-	53.3	-	32.9	-	14.9	-	-	-	-	-	85.1	68	-	-	-	-
Cascaded Randomized Smoothing 2023.10	90.1	-	40.5	-	25	-	16.2	-	8.8	-	-	-	79.6	64.5	-	-	-	-
Consistency 2023.10	89.5	-	60.7	-	-	-	-	-	-	-	-	-	84.9	51.7	-	-	-	-
Carlini et al. 2023.10	88.8	-	50.6	-	31	-	14.3	-	-	-	-	-	75.3	64.6	-	-	-	-
Diffusion Calibration 2023.10	88.6	-	53.3	-	32.5	-	15.3	-	-	-	-	-	75.1	66.4	-	-	-	-
Gaussian 2023.10	87.4	-	45.9	-	24.4	-	6	-	-	-	-	-	69.9	54.3	-	-	-	-
Carlini et al. 2023.10	87.3	-	32.5	-	21.9	-	11.8	-	6.4	-	-	-	48.6	53	-	-	-	-
Diffusion Calibration 2023.10	83.3	-	36.7	-	24.9	-	16.1	-	8.8	-	-	-	52.6	57.7	-	-	-	-
Consistency 2023.10	80.7	-	49.2	-	31.2	-	13.1	-	-	-	-	-	67.7	61.7	-	-	-	-
Gaussian 2023.10	75.6	-	27.9	-	16.9	-	9	-	4	-	-	-	42.2	42.6	-	-	-	-
LipConvnet-30 2022.10	72.08	-	-	-	-	-	-	-	-	-	-	-	-	-	62.84	51.99	41.94	-
LipConvnet-15 2022.10	71.89	-	-	-	-	-	-	-	-	-	-	-	-	-	62.37	52.24	42.71	-
LipConvnet-20 2022.10	71.86	-	-	-	-	-	-	-	-	-	-	-	-	-	62.86	52.24	42.39	-
LipConvnet-10 2022.10	71.82	-	-	-	-	-	-	-	-	-	-	-	-	-	62.6	51.76	42.31	-
LipConvnet-25 2022.10	71.64	-	-	-	-	-	-	-	-	-	-	-	-	-	62.67	52	42.34	-
LipConvnet-5 2022.10	71.52	-	-	-	-	-	-	-	-	-	-	-	-	-	61.25	50.66	40.31	-
LipConvnet-15 2022.10	71.15	-	-	-	-	-	-	-	-	-	-	-	-	-	61.65	51.78	42.53	-
LipConvnet-10 2022.10	71.1	-	-	-	-	-	-	-	-	-	-	-	-	-	60.81	50.61	41.03	-
LipConvnet-20 2022.10	70.95	-	-	-	-	-	-	-	-	-	-	-	-	-	61.72	51.78	42.01	-
LipConvnet-25 2022.10	70.58	-	-	-	-	-	-	-	-	-	-	-	-	-	61.4	51.46	42.05	-
LipConvnet-30 2022.10	70.37	-	-	-	-	-	-	-	-	-	-	-	-	-	60.74	51.23	41.81	-
LipConvnet-5 2022.10	69.67	-	-	-	-	-	-	-	-	-	-	-	-	-	59.28	48.02	38.3	-
Consistency 2023.10	65.8	-	31.1	-	22.3	-	14.6	-	8.7	-	-	-	43.1	51.3	-	-	-	-
Cohen et al. 2020.02	-	60	43	34	23	17	14	12	10	8	6	4	-	-	-	-	-	-
Salman et al. 2020.02	-	74	57	48	38	33	29	25	19	17	14	12	-	-	-	-	-	-
MaxUp 2020.02	-	74	57	49	40	35	31	27	22	19	17	15	-	-	-	-	-	-
Gaussian 2021.11	-	39.2	34	27.8	21.6	17.4	14	11.8	10	7.6	-	-	47.2	54.2	-	-	-	-
Stability training 2021.11	-	38.9	32.8	27	23.1	19.1	15.4	11.3	7.8	5.7	-	-	43.5	52.6	-	-	-	-
SmoothAdv 2021.11	-	44.9	39	33.6	28.5	23.7	19.4	15.4	12	8.7	-	-	50.8	66	-	-	-	-
MACER 2021.11	-	38.5	35.2	32.3	29.3	26.4	23.4	20.2	17.4	14.5	-	-	41.4	74.4	-	-	-	-
Consistency 2021.11	-	42.2	38.1	34.3	30	26.3	22.9	19.7	-	13.8	-	-	46.3	75.6	-	-	-	-
SmoothMix 2021.11	-	42.5	37.5	32.9	28.7	24.9	21.3	18.3	15.5	12.6	-	-	47.1	72.5	-	-	-	-
SmoothMix + One-step adversary 2021.11	-	41.5	37.5	33.8	30.2	26.7	23.4	20.2	17.2	14.7	-	-	45.1	77.3	-	-	-	-
Hu et al. (2023) 2024.10	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	69.5
Hu et al. (2023) 2024.10	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	52.2
CURE-Scratch 2024.10	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	71.2
CURE-Scratch 2024.10	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	56.6