Domain Generalization On Domainnet

评估指标

Average Accuracy

评测结果

各个模型在此基准测试上的表现结果

		Paper Title	Repository
PromptStyler (CLIP, ViT-L/14)	65.5	PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization
VDPG (CLIP, ViT-L/14)	65.2	Adapting to Distribution Shift by Visual Domain Prompt Generation
VL2V-SD (CLIP, ViT-B/16)	62.79	Leveraging Vision-Language Models for Improving Domain Generalization in Image Classification
MoA (OpenCLIP, ViT-B/16)	62.7	Domain Generalization Using Large Pretrained Models with Mixture-of-Adapters
CAR-FT (CLIP, ViT-B/16)	62.5	Context-Aware Robust Fine-Tuning	-
SIMPLE+	61.9	SIMPLE: Specialized Model-Sample Matching for Domain Generalization	-
GMDG (RegNetY-16GF, SWAD)	61.3	Rethinking Multi-domain Generalization with A General Learning Objective
Ensemble of Averages (RegNetY-16GF)	60.9	Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization
MIRO (RegNetY-16GF, SWAD)	60.7	Domain Generalization by Mutual-Information Regularization with Pre-trained Models
SPG (CLIP, ViT-B/16)	60.1	Soft Prompt Generation for Domain Generalization
VDPG (CLIP, ViT-B/16)	59.8	Adapting to Distribution Shift by Visual Domain Prompt Generation
UniDG + CORAL + ConvNeXt-B	59.5	Towards Unified and Effective Domain Generalization
PromptStyler (CLIP, ViT-B/16)	59.4	PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization
SEDGE+	54.7	Domain Generalization using Pretrained Models without Fine-tuning	-
GMDG (RegNetY-16GF)	54.6	Rethinking Multi-domain Generalization with A General Learning Objective
Ensemble of Averages (ResNeXt-50 32x4d)	54.6	Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization
Hybrid-SF-MoE	52.0	Sparse Mixture-of-Experts are Domain Generalizable Learners
CADG	51.6	CADG: A Model Based on Cross Attention for Domain Generalization	-
SPG (CLIP, ResNet-50)	50.1	Soft Prompt Generation for Domain Generalization
PromptStyler (CLIP, ResNet-50)	49.5	PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization

0 of 36 row(s) selected.