Domain Generalization on DomainNet

Metric: Average Accuracy (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	Average Accuracy▼	Extra Data	Paper	Date↕	Code
1	L2C (CLIP, ViT-L/14)	67.4	No	-	-	Code
2	PromptStyler (CLIP, ViT-L/14)	65.5	No	PromptStyler: Prompt-driven Style Generation for...	2023-07-27	Code
3	VDPG (CLIP, ViT-L/14)	65.2	No	Adapting to Distribution Shift by Visual Domain ...	2024-05-05	Code
4	VL2V-SD (CLIP, ViT-B/16)	62.79	No	Leveraging Vision-Language Models for Improving ...	2023-10-12	Code
5	MoA (OpenCLIP, ViT-B/16)	62.7	No	Domain Generalization Using Large Pretrained Mod...	2023-10-17	Code
6	CAR-FT (CLIP, ViT-B/16)	62.5	Yes	Context-Aware Robust Fine-Tuning	2022-11-29	-
7	SIMPLE+	61.9	No	-	-	Code
8	GMDG (RegNetY-16GF, SWAD)	61.3	No	Rethinking Multi-domain Generalization with A Ge...	2024-02-29	Code
9	L2C (CLIP, ViT-B/16)	61.2	No	-	-	Code
10	Ensemble of Averages (RegNetY-16GF)	60.9	Yes	Ensemble of Averages: Improving Model Selection ...	2021-10-21	Code
11	MIRO (RegNetY-16GF, SWAD)	60.7	Yes	Domain Generalization by Mutual-Information Regu...	2022-03-21	Code
12	SPG (CLIP, ViT-B/16)	60.1	No	Soft Prompt Generation for Domain Generalization	2024-04-30	Code
13	VDPG (CLIP, ViT-B/16)	59.8	No	Adapting to Distribution Shift by Visual Domain ...	2024-05-05	Code
14	UniDG + CORAL + ConvNeXt-B	59.5	No	Towards Unified and Effective Domain Generalizat...	2023-10-16	Code
15	PromptStyler (CLIP, ViT-B/16)	59.4	No	PromptStyler: Prompt-driven Style Generation for...	2023-07-27	Code
16	SEDGE+	54.7	No	Domain Generalization using Pretrained Models wi...	2022-03-09	-
17	Ensemble of Averages (ResNeXt-50 32x4d)	54.6	Yes	Ensemble of Averages: Improving Model Selection ...	2021-10-21	Code
18	GMDG (RegNetY-16GF)	54.6	No	Rethinking Multi-domain Generalization with A Ge...	2024-02-29	Code
19	Hybrid-SF-MoE	52	Yes	Sparse Mixture-of-Experts are Domain Generalizab...	2022-06-08	Code
20	CADG	51.6	Yes	CADG: A Model Based on Cross Attention for Domai...	2022-03-31	-
21	SPG (CLIP, ResNet-50)	50.1	No	Soft Prompt Generation for Domain Generalization	2024-04-30	Code
22	PromptStyler (CLIP, ResNet-50)	49.5	No	PromptStyler: Prompt-driven Style Generation for...	2023-07-27	Code
23	SIMPLE	49.2	No	-	-	Code
24	GMoE-S/16	48.7	No	Sparse Mixture-of-Experts are Domain Generalizab...	2022-06-08	Code
25	Ensemble of Averages (ResNet-50)	47.4	No	Ensemble of Averages: Improving Model Selection ...	2021-10-21	Code
26	GMDG (ResNet-50, SWAD)	47.3	No	Rethinking Multi-domain Generalization with A Ge...	2024-02-29	Code
27	MIRO (ResNet-50, SWAD)	47	No	Domain Generalization by Mutual-Information Regu...	2022-03-21	Code
28	DDG	46.93	No	Dynamic Domain Generalization	2022-05-27	Code
29	AdaClust (ResNet-50, SWAD)	46.7	No	Adaptive Methods for Aggregated Domain Generaliz...	2021-12-09	Code
30	SWAD (ResNet-50)	46.5	No	SWAD: Domain Generalization by Seeking Flat Minima	2021-02-17	Code
31	SEDGE	46.3	No	Domain Generalization using Pretrained Models wi...	2022-03-09	-
32	GMDG (ResNet-50)	44.6	No	Rethinking Multi-domain Generalization with A Ge...	2024-02-29	Code
33	Meta-DMoE (ResNet-50)	44.2	No	Meta-DMoE: Adapting to Domain Shift by Meta-Dist...	2022-10-08	Code
34	POEM	44	No	POEM: Polarization of Embeddings for Domain-Inva...	2023-05-22	Code
35	DMG (ResNet-50)	43.63	No	Learning to Balance Specificity and Invariance f...	2020-08-28	Code
36	MetaReg (ResNet-50)	43.62	No	Learning to Balance Specificity and Invariance f...	2020-08-28	Code
37	AdaClust (ResNet-50)	43.3	No	Adaptive Methods for Aggregated Domain Generaliz...	2021-12-09	Code
38	Fishr (ResNet-50)	41.8	No	Fishr: Invariant Gradient Variances for Out-of-D...	2021-09-07	Code

#1L2C (CLIP, ViT-L/14)
67.4
Average Accuracy
No paperCode
#2PromptStyler (CLIP, ViT-L/14)SOTA
65.5
Average Accuracy· 2023-07-27
PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization Code
#3VDPG (CLIP, ViT-L/14)
65.2
Average Accuracy· 2024-05-05
Adapting to Distribution Shift by Visual Domain Prompt Generation Code
#4VL2V-SD (CLIP, ViT-B/16)
62.79
Average Accuracy· 2023-10-12
Leveraging Vision-Language Models for Improving Domain Generalization in Image Classification Code
#5MoA (OpenCLIP, ViT-B/16)
62.7
Average Accuracy· 2023-10-17
Domain Generalization Using Large Pretrained Models with Mixture-of-Adapters Code
#6CAR-FT (CLIP, ViT-B/16)SOTA
62.5
Average Accuracy· Extra Data· 2022-11-29
Context-Aware Robust Fine-Tuning
#7SIMPLE+
61.9
Average Accuracy
No paperCode
#8GMDG (RegNetY-16GF, SWAD)
61.3
Average Accuracy· 2024-02-29
Rethinking Multi-domain Generalization with A General Learning Objective Code
#9L2C (CLIP, ViT-B/16)
61.2
Average Accuracy
No paperCode
#10Ensemble of Averages (RegNetY-16GF)SOTA
60.9
Average Accuracy· Extra Data· 2021-10-21
Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization Code
#11MIRO (RegNetY-16GF, SWAD)
60.7
Average Accuracy· Extra Data· 2022-03-21
Domain Generalization by Mutual-Information Regularization with Pre-trained Models Code
#12SPG (CLIP, ViT-B/16)
60.1
Average Accuracy· 2024-04-30
Soft Prompt Generation for Domain Generalization Code
#13VDPG (CLIP, ViT-B/16)
59.8
Average Accuracy· 2024-05-05
Adapting to Distribution Shift by Visual Domain Prompt Generation Code
#14UniDG + CORAL + ConvNeXt-B
59.5
Average Accuracy· 2023-10-16
Towards Unified and Effective Domain Generalization Code
#15PromptStyler (CLIP, ViT-B/16)
59.4
Average Accuracy· 2023-07-27
PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization Code
#16SEDGE+
54.7
Average Accuracy· 2022-03-09
Domain Generalization using Pretrained Models without Fine-tuning
#17Ensemble of Averages (ResNeXt-50 32x4d)
54.6
Average Accuracy· Extra Data· 2021-10-21
Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization Code
#18GMDG (RegNetY-16GF)
54.6
Average Accuracy· 2024-02-29
Rethinking Multi-domain Generalization with A General Learning Objective Code
#19Hybrid-SF-MoE
52
Average Accuracy· Extra Data· 2022-06-08
Sparse Mixture-of-Experts are Domain Generalizable Learners Code
#20CADG
51.6
Average Accuracy· Extra Data· 2022-03-31
CADG: A Model Based on Cross Attention for Domain Generalization
#21SPG (CLIP, ResNet-50)
50.1
Average Accuracy· 2024-04-30
Soft Prompt Generation for Domain Generalization Code
#22PromptStyler (CLIP, ResNet-50)
49.5
Average Accuracy· 2023-07-27
PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization Code
#23SIMPLE
49.2
Average Accuracy
No paperCode
#24GMoE-S/16
48.7
Average Accuracy· 2022-06-08
Sparse Mixture-of-Experts are Domain Generalizable Learners Code
#25Ensemble of Averages (ResNet-50)
47.4
Average Accuracy· 2021-10-21
Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization Code
#26GMDG (ResNet-50, SWAD)
47.3
Average Accuracy· 2024-02-29
Rethinking Multi-domain Generalization with A General Learning Objective Code
#27MIRO (ResNet-50, SWAD)
47
Average Accuracy· 2022-03-21
Domain Generalization by Mutual-Information Regularization with Pre-trained Models Code
#28DDG
46.93
Average Accuracy· 2022-05-27
Dynamic Domain Generalization Code
#29AdaClust (ResNet-50, SWAD)
46.7
Average Accuracy· 2021-12-09
Adaptive Methods for Aggregated Domain Generalization Code
#30SWAD (ResNet-50)SOTA
46.5
Average Accuracy· 2021-02-17
SWAD: Domain Generalization by Seeking Flat Minima Code
#31SEDGE
46.3
Average Accuracy· 2022-03-09
Domain Generalization using Pretrained Models without Fine-tuning
#32GMDG (ResNet-50)
44.6
Average Accuracy· 2024-02-29
Rethinking Multi-domain Generalization with A General Learning Objective Code
#33Meta-DMoE (ResNet-50)
44.2
Average Accuracy· 2022-10-08
Meta-DMoE: Adapting to Domain Shift by Meta-Distillation from Mixture-of-Experts Code
#34POEM
44
Average Accuracy· 2023-05-22
POEM: Polarization of Embeddings for Domain-Invariant Representations Code
#35DMG (ResNet-50)SOTA
43.63
Average Accuracy· 2020-08-28
Learning to Balance Specificity and Invariance for In and Out of Domain Generalization Code
#36MetaReg (ResNet-50)
43.62
Average Accuracy· 2020-08-28
Learning to Balance Specificity and Invariance for In and Out of Domain Generalization Code
#37AdaClust (ResNet-50)
43.3
Average Accuracy· 2021-12-09
Adaptive Methods for Aggregated Domain Generalization Code
#38Fishr (ResNet-50)
41.8
Average Accuracy· 2021-09-07
Fishr: Invariant Gradient Variances for Out-of-Distribution Generalization Code