# Image Classification on ImageNet C R Sketch A (test)

Source: [Image Classification on ImageNet C R Sketch A (test)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-c-r-sketch-a-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-17T09:53:32.867Z

Evaluation of Image Classification performance across the combined test sets of ImageNet-C (Corruption), ImageNet-R (Rendition), ImageNet-Sketch, and ImageNet-A (Adversarial) variants.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Image Classification |
| Dataset (including any stated split) | ImageNet C R Sketch A 1.0 (test) |
| Metric | Accuracy (ImageNet-C) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Accuracy (ImageNet-C)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-c-r-sketch-a-test.md?metric=Accuracy%20(ImageNet-C)>): Higher is better; unit: PERCENT.

- [Accuracy (ImageNet-R)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-c-r-sketch-a-test.md?metric=Accuracy%20(ImageNet-R)>): Higher is better; unit: PERCENT.

- [Accuracy (ImageNet-Sketch)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-c-r-sketch-a-test.md?metric=Accuracy%20(ImageNet-Sketch)>): Higher is better; unit: PERCENT.

- [Accuracy (ImageNet-A)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-c-r-sketch-a-test.md?metric=Accuracy%20(ImageNet-A)>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-c-r-sketch-a-test.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (ImageNet-C) | Result date | Report paper |
| --- | --- | --- | --- |
| EATA | 47.1 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| SAR | 46.8 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| ZOTTA | 44.9 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| DeYO | 34.9 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| CoTTA | 32.7 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| TENT | 31.4 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| NoAdapt | 31.3 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| T3A | 30.9 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| LAME | 30.5 | Not recorded | 69b8e03b0636e5bf4afbed59 |
| FOA | 24.5 | Not recorded | 69b8e03b0636e5bf4afbed59 |

### EATA

Accuracy (ImageNet-C): 47.1. Result ID: 69b9248d47f34ce924a242aa.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ca11e209131cea35123.md>)

Recorded evaluation configuration: {"Type":"BP-based","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### SAR

Accuracy (ImageNet-C): 46.8. Result ID: 69b9248d47f34ce924a242a9.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e961e209131cea40e77.md>)

Recorded evaluation configuration: {"Type":"BP-based","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### ZOTTA

Accuracy (ImageNet-C): 44.9. Result ID: 69b9248d47f34ce924a242af.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

Recorded evaluation configuration: {"Type":"BP-free","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### DeYO

Accuracy (ImageNet-C): 34.9. Result ID: 69b9248d47f34ce924a242ab.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72801e209131cea56d2d.md>)

[Implementation](<https://whitesnowdrop.github.io/DeYO/>)

Recorded evaluation configuration: {"Type":"BP-based","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### CoTTA

Accuracy (ImageNet-C): 32.7. Result ID: 69b9248d47f34ce924a242a8.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c191e209131cea32283.md>)

[Implementation](<https://qin.ee/cotta>)

Recorded evaluation configuration: {"Type":"BP-based","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### TENT

Accuracy (ImageNet-C): 31.4. Result ID: 69b9248d47f34ce924a242a7.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69941e209131cea22d2b.md>)

Recorded evaluation configuration: {"Type":"BP-based","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### NoAdapt

Accuracy (ImageNet-C): 31.3. Result ID: 69b9248d47f34ce924a242a6.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

Recorded evaluation configuration: {"Type":"/","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### T3A

Accuracy (ImageNet-C): 30.9. Result ID: 69b9248d47f34ce924a242ad.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

Recorded evaluation configuration: {"Type":"BP-free","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### LAME

Accuracy (ImageNet-C): 30.5. Result ID: 69b9248d47f34ce924a242ac.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c341e209131cea32ca1.md>)

[Implementation](<https://github.com/fiveai/LAME>)

Recorded evaluation configuration: {"Type":"BP-free","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

### FOA

Accuracy (ImageNet-C): 24.5. Result ID: 69b9248d47f34ce924a242ae.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e03b0636e5bf4afbed59.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73691e209131cea5c4c4.md>)

Recorded evaluation configuration: {"Type":"BP-free","Backbone":"ResNet50-GN"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
