# Image Classification on CIFAR-10 Unbalanced (val)

Source: [Image Classification on CIFAR-10 Unbalanced (val)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-unbalanced-val>)

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

Data updated: 2026-02-26T06:07:18.814Z

Evaluation of Image Classification performance on the validation split of the CIFAR-10 dataset configured with specific long-tail and two-level imbalance ratios.

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) | CIFAR-10 Unbalanced (val) |
| Metric | Top-1 Error (Long-tail, Ratio 100) |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 8 of 8 |
| 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

- [Top-1 Error (Long-tail, Ratio 100)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-unbalanced-val.md?metric=Top-1%20Error%20(Long-tail%2C%20Ratio%20100)>): Lower is better; unit: PERCENT.

- [Top-1 Error (Long-tail, Ratio 10)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-unbalanced-val.md?metric=Top-1%20Error%20(Long-tail%2C%20Ratio%2010)>): Lower is better; unit: PERCENT.

- [Top-1 Error (Two-level, Ratio 100)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-unbalanced-val.md?metric=Top-1%20Error%20(Two-level%2C%20Ratio%20100)>): Lower is better; unit: PERCENT.

- [Top-1 Error (Two-level, Ratio 10)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-unbalanced-val.md?metric=Top-1%20Error%20(Two-level%2C%20Ratio%2010)>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Error (Long-tail, Ratio 100) | Result date | Report paper |
| --- | --- | --- | --- |
| smDRAGON | 22.08 | Not recorded | 68fa68ac1e209131cea1daa1 |
| Focal (from \[11\]) | 25.43 | Not recorded | 68fa68ac1e209131cea1daa1 |
| LDAM | 26.65 | Not recorded | 68fa68ac1e209131cea1daa1 |
| ReWeight | 27.63 | Not recorded | 68fa68ac1e209131cea1daa1 |
| ReSample | 29.45 | Not recorded | 68fa68ac1e209131cea1daa1 |
| Focal (from \[27\]) | 29.62 | Not recorded | 68fa68ac1e209131cea1daa1 |
| CE | 29.64 | Not recorded | 68fa68ac1e209131cea1daa1 |
| CE\* (VE) | 29.81 | Not recorded | 68fa68ac1e209131cea1daa1 |

### smDRAGON

Top-1 Error (Long-tail, Ratio 100): 22.08. Result ID: 698d7e3f8b517eb43321ed41.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Implementation](<https://github.com/dvirsamuel/DRAGON>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### Focal (from \[11\])

Top-1 Error (Long-tail, Ratio 100): 25.43. Result ID: 698d7e3f8b517eb43321ed3c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa654f1e209131cea0988a.md>)

[Implementation](<https://github.com/richardaecn/class-balanced-loss>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### LDAM

Top-1 Error (Long-tail, Ratio 100): 26.65. Result ID: 698d7e3f8b517eb43321ed40.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66901e209131cea10e0f.md>)

[Implementation](<https://github.com/kaidic/LDAM-DRW>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### ReWeight

Top-1 Error (Long-tail, Ratio 100): 27.63. Result ID: 698d7e3f8b517eb43321ed3b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa654f1e209131cea0988a.md>)

[Implementation](<https://github.com/richardaecn/class-balanced-loss>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### ReSample

Top-1 Error (Long-tail, Ratio 100): 29.45. Result ID: 698d7e3f8b517eb43321ed3a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa654f1e209131cea0988a.md>)

[Implementation](<https://github.com/richardaecn/class-balanced-loss>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### Focal (from \[27\])

Top-1 Error (Long-tail, Ratio 100): 29.62. Result ID: 698d7e3f8b517eb43321ed3f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa640c1e209131cea02556.md>)

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

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### CE

Top-1 Error (Long-tail, Ratio 100): 29.64. Result ID: 698d7e3f8b517eb43321ed3d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66901e209131cea10e0f.md>)

[Implementation](<https://github.com/kaidic/LDAM-DRW>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

### CE\* (VE)

Top-1 Error (Long-tail, Ratio 100): 29.81. Result ID: 698d7e3f8b517eb43321ed3e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68ac1e209131cea1daa1.md>)

Recorded evaluation configuration: {"Vision-Only":true,"Backbone":"ResNet-32"}. This record does not assert independent reproduction.

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