# Image Classification on TinyImageNet Long-tail NID10.5 (test)

Source: [Image Classification on TinyImageNet Long-tail NID10.5 (test)](<https://www.sota2.com/research/sota/image-classification-on-tinyimagenet-long-tail-nid10-5-test>)

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

Data updated: 2026-05-14T09:35:54.600Z

Evaluation of Image Classification performance on the test split of the TinyImageNet Long-tail NID10.5 configuration, reporting accuracy metrics based on different rho thresholds (10, 50, 100).

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) | TinyImageNet Long-tail NID10.5 (test) |
| Metric | Accuracy (rho=10) |
| 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

- [Accuracy (rho=10)](<https://www.sota2.com/research/sota/image-classification-on-tinyimagenet-long-tail-nid10-5-test.md?metric=Accuracy%20(rho%3D10)>): Higher is better; unit: PERCENT.

- [Accuracy (rho=50)](<https://www.sota2.com/research/sota/image-classification-on-tinyimagenet-long-tail-nid10-5-test.md?metric=Accuracy%20(rho%3D50)>): Higher is better; unit: PERCENT.

- [Accuracy (rho=100)](<https://www.sota2.com/research/sota/image-classification-on-tinyimagenet-long-tail-nid10-5-test.md?metric=Accuracy%20(rho%3D100)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (rho=10) | Result date | Report paper |
| --- | --- | --- | --- |
| FedHPro | 37.72 | Not recorded | 6a05574edbc836b87d828fae |
| FedSA | 36.58 | Not recorded | 6a05574edbc836b87d828fae |
| FedRCL | 36.1 | Not recorded | 6a05574edbc836b87d828fae |
| FedGMKD | 35.74 | Not recorded | 6a05574edbc836b87d828fae |
| FedTGP | 35.51 | Not recorded | 6a05574edbc836b87d828fae |
| MOON | 35.29 | Not recorded | 6a05574edbc836b87d828fae |
| FedProx | 34.93 | Not recorded | 6a05574edbc836b87d828fae |
| FedAvg | 34.89 | Not recorded | 6a05574edbc836b87d828fae |
| FedProto | 34.4 | Not recorded | 6a05574edbc836b87d828fae |

### FedHPro

Accuracy (rho=10): 37.72. Result ID: 6a05960647f34ce924ecec1e.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"False","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedSA

Accuracy (rho=10): 36.58. Result ID: 6a05960647f34ce924ecec1d.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78751e209131cea786d2.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"True","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedRCL

Accuracy (rho=10): 36.1. Result ID: 6a05960647f34ce924ecec1c.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73611e209131cea5c107.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"False","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedGMKD

Accuracy (rho=10): 35.74. Result ID: 6a05960647f34ce924ecec1b.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"False","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedTGP

Accuracy (rho=10): 35.51. Result ID: 6a05960647f34ce924ecec1a.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71b41e209131cea52a38.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"True","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### MOON

Accuracy (rho=10): 35.29. Result ID: 6a05960647f34ce924ecec18.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa699a1e209131cea22fca.md>)

[Implementation](<https://github.com/QinbinLi/MOON>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"False","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedProx

Accuracy (rho=10): 34.93. Result ID: 6a05960647f34ce924ecec17.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67701e209131cea163fe.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"False","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedAvg

Accuracy (rho=10): 34.89. Result ID: 6a05960647f34ce924ecec16.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e761e209131cea40072.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"False","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

### FedProto

Accuracy (rho=10): 34.4. Result ID: 6a05960647f34ce924ecec19.

[Reporting paper](<https://www.sota2.com/research/paper/6a05574edbc836b87d828fae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea3152d.md>)

Recorded evaluation configuration: {"Prototype and Model Exchange":"True","Quantity Skew configuration":"NID10.5"}. This record does not assert independent reproduction.

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