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

Domain Generalization on DomainBed (out-of-domain)

86.5VLCS Accuracy

BOA

Updated 23d ago

Evaluation Results

Method
BOA 2026.05	86.5	97.4	86	60.3	60.2	78.1
SAGE 2026.05	85.5	97.4	87.1	61.1	63.4	78.9
GGA 2026.05	84.1	97.2	85.7	58.8	59.8	77.1
GAM 2026.05	83.6	96.4	85.5	55.3	59.5	76.1
IRM 2026.05	82.9	96.1	83.2	56.7	59.1	75.6
SAM 2026.05	82.9	96.6	85.4	56.2	59.8	76.2
GSAM 2026.05	82.9	96.6	85.6	55.4	59.8	76.1
SAGM 2026.05	82.8	96.8	85.2	58	76.4	76.4
FISH 2026.05	82.7	96.9	85	58	59.1	75
DISAM 2026.05	82.7	97.1	85.4	57.3	59.8	76.5
CORAL 2026.05	82.6	96.4	83.8	57.5	59.8	76
CDANN 2026.05	82.4	96.5	82.9	55.6	58.4	75.2
MMD 2026.05	82.3	95.8	83.6	57.4	59.9	75.8
IIB 2026.05	82.3	96.5	84.2	58.2	58.6	76
RSC 2026.05	82.2	96.5	83.2	58.2	59	75.8
ERM 2026.05	81.9	95.9	84.1	56.1	59.5	75.5
DANN 2026.05	81.8	96.3	83	56	58.4	75.1
MASS 2025.12	81.1	88.7	73.8	47.5	-	72.8
DART 2023.02	80.3	88.9	71.9	51.3	47.2	67.9
GMoE 2025.12	80.2	88.1	74.2	48.5	-	72.8
GMoE 2025.12	80	88.2	74.2	47.2	-	72.4
GMoE 2025.12	79.8	88.2	73.5	47.8	-	72.3
DynMoE 2025.12	79.4	88.4	73.6	46.6	-	72
SWAD 2023.02	79.1	88.1	70.6	50	46.5	66.9
SWAD 2022.12	79.1	88.1	70.6	50	46.5	66.9
Ensemble* of inter-training 2022.12	79	89.2	72.7	51.1	47.2	67.8
CORAL 2023.02	78.8	86.2	68.7	47.6	41.5	64.5
CORAL 2022.12	78.8	86.2	68.7	47.6	41.5	64.6
Model ratatouille 2022.12	78.7	90.5	73.4	49.2	47.7	67.9
DANN 2023.02	78.6	83.6	65.9	46.7	38.3	62.6
Model soups+ (DiWA runs) 2022.12	78.6	89	72.8	51.9	47.7	68
IRM 2023.02	78.5	83.5	64.3	47.6	33.9	61.6
DART 2023.02	78.5	87.3	70.1	48.7	45.8	66.1
Deep ensembles* 2022.12	78.5	87.6	70.8	49.2	47.7	66.8
Ensemble* (DiWA runs) 2022.12	78.5	88.1	71.7	50.8	47	67.2
Model soups (DiWA runs) 2022.12	78.5	88	71.5	51.6	47.7	67.5
Fusing 2022.12	78.5	88	71.5	46.7	44.4	65.8
Model ratatouille 2022.12	78.5	89.5	73.1	51.8	47.5	68.1
Model soups (DiWA runs) 2022.12	78.4	88.7	72.1	51.4	47.4	67.6
VREX 2023.02	78.3	84.9	66.4	46.4	33.6	61.9
Model ratatouille† 2022.12	78.3	89.8	73.5	52	47.7	68.3
MA 2022.12	78.2	87.5	70.6	50.3	46	66.5
Vanilla fine-tuning (DiWA runs) 2022.12	78.1	85.9	69.4	50.4	44.3	65.6
SagNet 2023.02	77.8	86.3	68.1	48.6	40.3	64.2
Inter-training 2022.12	77.7	89	69.9	46.7	44.5	65.6
ARM 2023.02	77.6	85.1	64.8	45.5	35.5	61.7
ERM 2023.02	77.5	85.5	66.5	46.1	40.9	63.3
MMD 2023.02	77.5	84.6	66.3	42.2	23.4	58.8
CDANN 2023.02	77.5	82.6	65.8	45.8	38.3	62
Vanilla fine-tuning 2022.12	77.5	85.5	66.5	46.1	40.9	63.3
Mixup 2023.02	77.4	84.6	68.1	47.9	39.2	63.4
MLDG 2023.02	77.2	84.9	66.8	47.7	41.2	63.6
MTL 2023.02	77.2	84.6	66.4	45.6	40.6	62.9
RSC 2023.02	77.1	85.2	65.5	46.6	38.9	62.7
GroupDRO 2023.02	76.7	84.4	66	43.2	33.3	60.7