# Image Classification on CIFAR-100 (Parameterized Robustness Metrics)

Source: [Image Classification on CIFAR-100 (Parameterized Robustness Metrics)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics>)

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

Data updated: 2026-05-11T09:50:28.182Z

Evaluation of Image Classification performance on the CIFAR-100 dataset, reporting accuracy under various parameterized conditions defined by M (sample size proxy) and gamma\_u (distribution skew factor).

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-100 |
| Metric | Accuracy (Uniform, M=24600, γu=1) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 12 of 12 |
| 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 (Uniform, M=24600, γu=1)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Uniform%2C%20M%3D24600%2C%20%CE%B3u%3D1)>): Higher is better; unit: PERCENT.

- [Accuracy (Long-tailed, M=24600, γu=5)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Long-tailed%2C%20M%3D24600%2C%20%CE%B3u%3D5)>): Higher is better; unit: PERCENT.

- [Accuracy (Arbitrary, M=24600, γu=5)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Arbitrary%2C%20M%3D24600%2C%20%CE%B3u%3D5)>): Higher is better; unit: PERCENT.

- [Accuracy (Uniform, M=19400, γu=1)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Uniform%2C%20M%3D19400%2C%20%CE%B3u%3D1)>): Higher is better; unit: PERCENT.

- [Accuracy (Long-tailed, M=19400, γu=10)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Long-tailed%2C%20M%3D19400%2C%20%CE%B3u%3D10)>): Higher is better; unit: PERCENT.

- [Accuracy (Arbitrary, M=19400, γu=10)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Arbitrary%2C%20M%3D19400%2C%20%CE%B3u%3D10)>): Higher is better; unit: PERCENT.

- [Accuracy (Uniform, M=17100, γu=1)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Uniform%2C%20M%3D17100%2C%20%CE%B3u%3D1)>): Higher is better; unit: PERCENT.

- [Accuracy (Long-tailed, M=17100, γu=15)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Long-tailed%2C%20M%3D17100%2C%20%CE%B3u%3D15)>): Higher is better; unit: PERCENT.

- [Accuracy (Arbitrary, M=17100, γu=15)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Accuracy%20(Arbitrary%2C%20M%3D17100%2C%20%CE%B3u%3D15)>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-parameterized-robustness-metrics.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (Uniform, M=24600, γu=1) | Result date | Report paper |
| --- | --- | --- | --- |
| CE | 70.93 | Not recorded | 6a0162b1dbc836b87d82327c |
| LA | 70.83 | Not recorded | 6a0162b1dbc836b87d82327c |
| SAGE | 41 | Not recorded | 6a0162b1dbc836b87d82327c |
| CGMatch | 39.44 | Not recorded | 6a0162b1dbc836b87d82327c |
| ACR | 38.81 | Not recorded | 6a0162b1dbc836b87d82327c |
| CPG | 38.09 | Not recorded | 6a0162b1dbc836b87d82327c |
| FixMatch | 37.57 | Not recorded | 6a0162b1dbc836b87d82327c |
| FreeMatch | 36.53 | Not recorded | 6a0162b1dbc836b87d82327c |
| SoftMatch | 36.2 | Not recorded | 6a0162b1dbc836b87d82327c |
| Meta-Expert | 29.68 | Not recorded | 6a0162b1dbc836b87d82327c |
| CDMAD | 26.35 | Not recorded | 6a0162b1dbc836b87d82327c |
| SimPro | 19.1 | Not recorded | 6a0162b1dbc836b87d82327c |

### CE

Accuracy (Uniform, M=24600, γu=1): 70.93. Result ID: 6a01a66247f34ce924e3f91e.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

Recorded evaluation configuration: {"Scenario":"SL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### LA

Accuracy (Uniform, M=24600, γu=1): 70.83. Result ID: 6a01a66247f34ce924e3f91f.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

Recorded evaluation configuration: {"Scenario":"SL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### SAGE

Accuracy (Uniform, M=24600, γu=1): 41. Result ID: 6a01a66247f34ce924e3f929.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

Recorded evaluation configuration: {"Scenario":"LTSSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### CGMatch

Accuracy (Uniform, M=24600, γu=1): 39.44. Result ID: 6a01a66247f34ce924e3f923.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773b1e209131cea71df2.md>)

[Implementation](<https://github.com/BoCheng-96/CGMatch>)

Recorded evaluation configuration: {"Scenario":"SSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### ACR

Accuracy (Uniform, M=24600, γu=1): 38.81. Result ID: 6a01a66247f34ce924e3f924.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

Recorded evaluation configuration: {"Scenario":"LTSSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### CPG

Accuracy (Uniform, M=24600, γu=1): 38.09. Result ID: 6a01a66247f34ce924e3f928.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c101e209131cea8c36f.md>)

Recorded evaluation configuration: {"Scenario":"LTSSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### FixMatch

Accuracy (Uniform, M=24600, γu=1): 37.57. Result ID: 6a01a66247f34ce924e3f920.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68d61e209131cea1e896.md>)

Recorded evaluation configuration: {"Scenario":"SSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### FreeMatch

Accuracy (Uniform, M=24600, γu=1): 36.53. Result ID: 6a01a66247f34ce924e3f921.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e711e209131cea3fd15.md>)

[Implementation](<https://github.com/microsoft/Semi-supervised-learning>)

Recorded evaluation configuration: {"Scenario":"SSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### SoftMatch

Accuracy (Uniform, M=24600, γu=1): 36.2. Result ID: 6a01a66247f34ce924e3f922.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ec31e209131cea41dad.md>)

Recorded evaluation configuration: {"Scenario":"SSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### Meta-Expert

Accuracy (Uniform, M=24600, γu=1): 29.68. Result ID: 6a01a66247f34ce924e3f927.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a8d1e209131cea83dc2.md>)

Recorded evaluation configuration: {"Scenario":"LTSSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### CDMAD

Accuracy (Uniform, M=24600, γu=1): 26.35. Result ID: 6a01a66247f34ce924e3f926.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73551e209131cea5be23.md>)

[Implementation](<https://github.com/LeeHyuck/CDMAD>)

Recorded evaluation configuration: {"Scenario":"LTSSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

### SimPro

Accuracy (Uniform, M=24600, γu=1): 19.1. Result ID: 6a01a66247f34ce924e3f925.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162b1dbc836b87d82327c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74511e209131cea6108d.md>)

Recorded evaluation configuration: {"Scenario":"LTSSL","Labeled Samples (N)":"400","Labeled Distribution Ratio (γl)":"1"}. This record does not assert independent reproduction.

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