# Image Classification on Multi-domain MNIST (train)

Source: [Image Classification on Multi-domain MNIST (train)](<https://www.sota2.com/research/sota/image-classification-on-multi-domain-mnist-train>)

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

Data updated: 2026-02-26T05:34:12.773Z

Evaluation of Image Classification performance on the training split of the Multi-domain MNIST dataset, reporting accuracy under various corruption conditions (clean, noise, blur, contrast).

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) | Multi-domain MNIST (train) |
| Metric | Accuracy (Clean) |
| 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 (Clean)](<https://www.sota2.com/research/sota/image-classification-on-multi-domain-mnist-train.md?metric=Accuracy%20(Clean)>): Higher is better; unit: PERCENT.

- [Accuracy (Noise)](<https://www.sota2.com/research/sota/image-classification-on-multi-domain-mnist-train.md?metric=Accuracy%20(Noise)>): Higher is better; unit: PERCENT.

- [Accuracy (Blur)](<https://www.sota2.com/research/sota/image-classification-on-multi-domain-mnist-train.md?metric=Accuracy%20(Blur)>): Higher is better; unit: PERCENT.

- [Accuracy (Contrast)](<https://www.sota2.com/research/sota/image-classification-on-multi-domain-mnist-train.md?metric=Accuracy%20(Contrast)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (Clean) | Result date | Report paper |
| --- | --- | --- | --- |
| IRM | 99.47 | Not recorded | 697ae97aa2b28e761d87cfaf |
| GroupDRO | 99.1 | Not recorded | 697ae97aa2b28e761d87cfaf |
| MLDG | 98.9 | Not recorded | 697ae97aa2b28e761d87cfaf |
| baseline (ERM) | 98.4 | Not recorded | 697ae97aa2b28e761d87cfaf |
| MMD | 98.38 | Not recorded | 697ae97aa2b28e761d87cfaf |
| PNS | 96.78 | Not recorded | 697ae97aa2b28e761d87cfaf |
| PNS | 96.77 | Not recorded | 697ae97aa2b28e761d87cfaf |
| PNS | 96.66 | Not recorded | 697ae97aa2b28e761d87cfaf |
| CDANN | 96.23 | Not recorded | 697ae97aa2b28e761d87cfaf |
| CaSN(IRM) | 95.58 | Not recorded | 697ae97aa2b28e761d87cfaf |

### IRM

Accuracy (Clean): 99.47. Result ID: 699908c65698665ec46e4660.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa673c1e209131cea15133.md>)

[Implementation](<https://github.com/facebookresearch/InvariantRiskMinimization>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### GroupDRO

Accuracy (Clean): 99.1. Result ID: 699908c65698665ec46e465d.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### MLDG

Accuracy (Clean): 98.9. Result ID: 699908c65698665ec46e4662.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### baseline (ERM)

Accuracy (Clean): 98.4. Result ID: 699908c65698665ec46e4663.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### MMD

Accuracy (Clean): 98.38. Result ID: 699908c65698665ec46e465e.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### PNS

Accuracy (Clean): 96.78. Result ID: 699908c65698665ec46e4665.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: {"segment selection":"Top-3"}. This record does not assert independent reproduction.

### PNS

Accuracy (Clean): 96.77. Result ID: 699908c65698665ec46e4666.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: {"segment selection":"Top-4"}. This record does not assert independent reproduction.

### PNS

Accuracy (Clean): 96.66. Result ID: 699908c65698665ec46e4664.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: {"segment selection":"Top-2"}. This record does not assert independent reproduction.

### CDANN

Accuracy (Clean): 96.23. Result ID: 699908c65698665ec46e465f.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### CaSN(IRM)

Accuracy (Clean): 95.58. Result ID: 699908c65698665ec46e4661.

[Reporting paper](<https://www.sota2.com/research/paper/697ae97aa2b28e761d87cfaf.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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