# Image Classification on Long-tailed ImageNet (val)

Source: [Image Classification on Long-tailed ImageNet (val)](<https://www.sota2.com/research/sota/image-classification-on-long-tailed-imagenet-val>)

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

Data updated: 2026-02-26T05:26:51.424Z

Evaluation of Image Classification performance on the validation split of the Long-tailed ImageNet dataset, focusing on accuracy across different class frequency buckets (Many, Medium, Few).

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) | Long-tailed ImageNet (val) |
| Metric | Acc (Many) |
| Unit | PERCENT |
| Direction | Higher 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

- [Acc (Many)](<https://www.sota2.com/research/sota/image-classification-on-long-tailed-imagenet-val.md?metric=Acc%20(Many)>): Higher is better; unit: PERCENT.

- [Acc (Medium)](<https://www.sota2.com/research/sota/image-classification-on-long-tailed-imagenet-val.md?metric=Acc%20(Medium)>): Higher is better; unit: PERCENT.

- [Acc (Few)](<https://www.sota2.com/research/sota/image-classification-on-long-tailed-imagenet-val.md?metric=Acc%20(Few)>): Higher is better; unit: PERCENT.

- [Accuracy](<https://www.sota2.com/research/sota/image-classification-on-long-tailed-imagenet-val.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Acc (Many) | Result date | Report paper |
| --- | --- | --- | --- |
| CE | 64 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| DRO-LT | 64 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| DRO-LT | 63.8 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| DRO-LT | 62.6 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| SSL | 61.4 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| DRO-LT | 61 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| smDragon | 59.7 | Not recorded | 68fa6b2d1e209131cea2c9d1 |
| T-norm | 56.6 | Not recorded | 68fa6b2d1e209131cea2c9d1 |

### CE

Acc (Many): 64. Result ID: 698e695fffc3901de09d60e1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

Recorded evaluation configuration: {"reproduced":"true"}. This record does not assert independent reproduction.

### DRO-LT

Acc (Many): 64. Result ID: 698e695fffc3901de09d60e8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

Recorded evaluation configuration: {"epsilon":"Learnable"}. This record does not assert independent reproduction.

### DRO-LT

Acc (Many): 63.8. Result ID: 698e695fffc3901de09d60e7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

Recorded evaluation configuration: {"epsilon":"e\[n\]"}. This record does not assert independent reproduction.

### DRO-LT

Acc (Many): 62.6. Result ID: 698e695fffc3901de09d60e6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

Recorded evaluation configuration: {"epsilon":"Shared"}. This record does not assert independent reproduction.

### SSL

Acc (Many): 61.4. Result ID: 698e695fffc3901de09d60e4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa697f1e209131cea22756.md>)

Recorded evaluation configuration: {"reproduced":"true"}. This record does not assert independent reproduction.

### DRO-LT

Acc (Many): 61. Result ID: 698e695fffc3901de09d60e5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

Recorded evaluation configuration: {"epsilon":"0 (ERM)"}. This record does not assert independent reproduction.

### smDragon

Acc (Many): 59.7. Result ID: 698e695fffc3901de09d60e3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

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

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

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

### T-norm

Acc (Many): 56.6. Result ID: 698e695fffc3901de09d60e2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2d1e209131cea2c9d1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa671b1e209131cea144a9.md>)

[Implementation](<https://github.com/facebookresearch/classifier-balancing>)

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.
