ML Research Wiki / Benchmarks / Unsupervised Domain Adaptation / ImageNet-R

ImageNet-R

Unsupervised Domain Adaptation Benchmark

Performance Over Time

📊 Showing 8 results | 📏 Metric: Top 1 Error

Top Performing Models

Rank	Model	Paper	Top 1 Error	Date	Code
1	ResNet50, BatchNorm adaptation	Improving robustness against common corruptions by covariate shift adaptation	59.90	2020-06-30	📦 bethgelab/robustness 📦 Claydon-Wang/OFTTA
2	ResNet50 + ENT	If your data distribution shifts, use self-learning	56.10	2021-04-27	📦 bethgelab/robustness
3	ResNet50 + RPL	If your data distribution shifts, use self-learning	54.10	2021-04-27	📦 bethgelab/robustness
4	ResNet50+DeepAug+Augmix, BatchNorm adaptation	Improving robustness against common corruptions by covariate shift adaptation	48.90	2020-06-30	📦 bethgelab/robustness 📦 Claydon-Wang/OFTTA
5	ResNeXt101+DeepAug+AugMix, BatchNorm Adaptation,	Improving robustness against common corruptions by covariate shift adaptation	44.00	2020-06-30	📦 bethgelab/robustness 📦 Claydon-Wang/OFTTA
6	EfficientNet-L2 Noisy Student + ENT 📚	If your data distribution shifts, use self-learning	19.70	2021-04-27	📦 bethgelab/robustness
7	EfficientNet-L2 Noisy Student + RPL 📚	If your data distribution shifts, use self-learning	17.40	2021-04-27	📦 bethgelab/robustness
8	Model soups (ViT-G/14) 📚	Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time	4.54	2022-03-10	📦 mlfoundations/model-soups 📦 Burf/ModelSoups 📦 facebookresearch/ModelRatatouille

All Papers (8)

Improving robustness against common corruptions by covariate shift adaptation

2020

ResNet50, BatchNorm adaptation

bethgelab/robustness Claydon-Wang/OFTTA

If your data distribution shifts, use self-learning

2021

ResNet50 + ENT

bethgelab/robustness

If your data distribution shifts, use self-learning

2021

ResNet50 + RPL

bethgelab/robustness

Improving robustness against common corruptions by covariate shift adaptation

2020

ResNet50+DeepAug+Augmix, BatchNorm adaptation

bethgelab/robustness Claydon-Wang/OFTTA

Improving robustness against common corruptions by covariate shift adaptation

2020

ResNeXt101+DeepAug+AugMix, BatchNorm Adaptation,

bethgelab/robustness Claydon-Wang/OFTTA

If your data distribution shifts, use self-learning

2021

EfficientNet-L2 Noisy Student + ENT

bethgelab/robustness

If your data distribution shifts, use self-learning

2021

EfficientNet-L2 Noisy Student + RPL

bethgelab/robustness

Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time

2022

Model soups (ViT-G/14)

mlfoundations/model-soups Burf/ModelSoups

ImageNet-R

Performance Over Time

Edit Benchmark Results

Edit Result

Top Performing Models

All Papers (8)

Improving robustness against common corruptions by covariate shift adaptation

If your data distribution shifts, use self-learning

If your data distribution shifts, use self-learning

Improving robustness against common corruptions by covariate shift adaptation

Improving robustness against common corruptions by covariate shift adaptation

If your data distribution shifts, use self-learning

If your data distribution shifts, use self-learning

Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time

Model	Paper	Top 1 Error	Date
ResNet50, BatchNorm adaptation	Improving robustness against common corruptions b…	59.90	2020-06-30
ResNet50 + ENT	If your data distribution shifts, use self-learni…	56.10	2021-04-27
ResNet50 + RPL	If your data distribution shifts, use self-learni…	54.10	2021-04-27
ResNet50+DeepAug+Augmix, BatchNorm adaptation	Improving robustness against common corruptions b…	48.90	2020-06-30
ResNeXt101+DeepAug+AugMix, BatchNorm Adaptation,	Improving robustness against common corruptions b…	44.00	2020-06-30
EfficientNet-L2 Noisy Student + ENT	If your data distribution shifts, use self-learni…	19.70	2021-04-27
EfficientNet-L2 Noisy Student + RPL	If your data distribution shifts, use self-learni…	17.40	2021-04-27
Model soups (ViT-G/14)	Model soups: averaging weights of multiple fine-t…	4.54	2022-03-10