# Image Classification on ImageNet Robustness Variants (V2, A, R, S)

Source: [Image Classification on ImageNet Robustness Variants (V2, A, R, S)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-robustness-variants-v2-a-r-s>)

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

Data updated: 2026-02-26T05:35:14.746Z

Evaluation of Image Classification performance specifically across the four primary ImageNet robustness variants: ImageNet-V2, ImageNet-S, ImageNet-A, and ImageNet-R.

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) | 4 ImageNet Variants (ImgNet-V2, ImgNet-S, ImgNet-A, ImgNet-R) |
| Metric | Average Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Average Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-robustness-variants-v2-a-r-s.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Skip Tuning | 61.2 | Not recorded | 68fa77741e209131cea7337d |
| ProSRC | 60.65 | Not recorded | 68fa77741e209131cea7337d |
| MaPLe | 60.26 | Not recorded | 68fa77741e209131cea7337d |
| KgCoOp | 60.11 | Not recorded | 68fa77741e209131cea7337d |
| CoPro | 60.02 | Not recorded | 68fa77741e209131cea7337d |
| CoCoOp | 59.9 | Not recorded | 68fa77741e209131cea7337d |
| CoOp | 59.28 | Not recorded | 68fa77741e209131cea7337d |
| ProGrad | 59.07 | Not recorded | 68fa77741e209131cea7337d |
| DePT | 58.97 | Not recorded | 68fa77741e209131cea7337d |

### Skip Tuning

Average Accuracy: 61.2. Result ID: 69948bb53a6c3148491de966.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Implementation](<https://github.com/Koorye/SkipTuning>)

Recorded evaluation configuration: {"Time (s)":"1139","Mem. (M)":"656"}. This record does not assert independent reproduction.

### ProSRC

Average Accuracy: 60.65. Result ID: 69948bb53a6c3148491de963.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa705e1e209131cea4aa58.md>)

Recorded evaluation configuration: {"Time (s)":"50091","Mem. (M)":"14107"}. This record does not assert independent reproduction.

### MaPLe

Average Accuracy: 60.26. Result ID: 69948bb53a6c3148491de962.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef21e209131cea4311c.md>)

[Implementation](<https://github.com/muzairkhattak/multimodal-prompt-learning>)

Recorded evaluation configuration: {"Time (s)":"4942","Mem. (M)":"13786"}. This record does not assert independent reproduction.

### KgCoOp

Average Accuracy: 60.11. Result ID: 69948bb53a6c3148491de961.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee91e209131cea42d7a.md>)

Recorded evaluation configuration: {"Time (s)":"31636","Mem. (M)":"15254"}. This record does not assert independent reproduction.

### CoPro

Average Accuracy: 60.02. Result ID: 69948bb53a6c3148491de965.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74641e209131cea616f7.md>)

[Implementation](<https://github.com/ShuvenduRoy/CoPrompt>)

Recorded evaluation configuration: {"Time (s)":"19161","Mem. (M)":"14131"}. This record does not assert independent reproduction.

### CoCoOp

Average Accuracy: 59.9. Result ID: 69948bb53a6c3148491de95f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d8c1e209131cea3a831.md>)

[Implementation](<https://github.com/KaiyangZhou/CoOp>)

Recorded evaluation configuration: {"Time (s)":"93917","Mem. (M)":"13622"}. This record does not assert independent reproduction.

### CoOp

Average Accuracy: 59.28. Result ID: 69948bb53a6c3148491de95e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d8b1e209131cea3a814.md>)

[Implementation](<https://github.com/KaiyangZhou/CoOp>)

Recorded evaluation configuration: {"Time (s)":"31632","Mem. (M)":"15412"}. This record does not assert independent reproduction.

### ProGrad

Average Accuracy: 59.07. Result ID: 69948bb53a6c3148491de960.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a981e209131cea841f7.md>)

Recorded evaluation configuration: {"Time (s)":"56223","Mem. (M)":"15412"}. This record does not assert independent reproduction.

### DePT

Average Accuracy: 58.97. Result ID: 69948bb53a6c3148491de964.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77741e209131cea7337d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea5705c.md>)

[Implementation](<https://github.com/Koorye/DePT>)

Recorded evaluation configuration: {"Time (s)":"22796","Mem. (M)":"13783"}. This record does not assert independent reproduction.

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