# Image Classification on CIFAR-100 standard (test) (10%, 30%, 40% Noise)

Source: [Image Classification on CIFAR-100 standard (test) (10%, 30%, 40% Noise)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-standard-test-10-percent-30-percent-40-percent-noise>)

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

Data updated: 2026-02-26T05:34:00.315Z

Evaluation of Image Classification performance on the CIFAR-100 test set under varying levels of input noise (10%, 30%, and 40%).

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 standard (test) |
| Metric | Accuracy (10% Noise) |
| 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 (10% Noise)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-standard-test-10-percent-30-percent-40-percent-noise.md?metric=Accuracy%20(10%25%20Noise)>): Higher is better; unit: PERCENT.

- [Accuracy (30% Noise)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-standard-test-10-percent-30-percent-40-percent-noise.md?metric=Accuracy%20(30%25%20Noise)>): Higher is better; unit: PERCENT.

- [Accuracy (40% Noise)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-standard-test-10-percent-30-percent-40-percent-noise.md?metric=Accuracy%20(40%25%20Noise)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (10% Noise) | Result date | Report paper |
| --- | --- | --- | --- |
| UNICON | 78.2 | Not recorded | 68fa6c691e209131cea33eab |
| MOIT | 77.4 | Not recorded | 68fa6c691e209131cea33eab |
| ELR\* | 77.3 | Not recorded | 68fa6c691e209131cea33eab |
| PCIL | 76 | Not recorded | 68fa6c691e209131cea33eab |
| M-Up | 72.4 | Not recorded | 68fa6c691e209131cea33eab |
| JPL | 72 | Not recorded | 68fa6c691e209131cea33eab |
| DMix\* | 71.6 | Not recorded | 68fa6c691e209131cea33eab |
| CE | 68.1 | Not recorded | 68fa6c691e209131cea33eab |
| LDMI | 68.1 | Not recorded | 68fa6c691e209131cea33eab |

### UNICON

Accuracy (10% Noise): 78.2. Result ID: 698f12043a6c314849f57988.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Implementation](<https://github.com/nazmul-karim170/UNICON-Noisy-Label>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

### MOIT

Accuracy (10% Noise): 77.4. Result ID: 698f12043a6c314849f57987.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69951e209131cea22d59.md>)

[Implementation](<https://git.io/JI40X>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

### ELR\*

Accuracy (10% Noise): 77.3. Result ID: 698f12043a6c314849f57986.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68a01e209131cea1d383.md>)

[Implementation](<https://github.com/shengliu66/ELR>)

Recorded evaluation configuration: {"noise\_type":"asymmetric","implementation":"re-run by authors"}. This record does not assert independent reproduction.

### PCIL

Accuracy (10% Noise): 76. Result ID: 698f12043a6c314849f57984.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65801e209131cea0a8aa.md>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

### M-Up

Accuracy (10% Noise): 72.4. Result ID: 698f12043a6c314849f57982.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa644e1e209131cea03e44.md>)

[Implementation](<https://github.com/facebookresearch/mixup-cifar10>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

### JPL

Accuracy (10% Noise): 72. Result ID: 698f12043a6c314849f57983.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b71e209131cea23ac4.md>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

### DMix\*

Accuracy (10% Noise): 71.6. Result ID: 698f12043a6c314849f57985.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa671c1e209131cea144cf.md>)

[Implementation](<https://github.com/LiJunnan1992/DivideMix>)

Recorded evaluation configuration: {"noise\_type":"asymmetric","implementation":"re-run by authors"}. This record does not assert independent reproduction.

### CE

Accuracy (10% Noise): 68.1. Result ID: 698f12043a6c314849f57980.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

### LDMI

Accuracy (10% Noise): 68.1. Result ID: 698f12043a6c314849f57981.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c691e209131cea33eab.md>)

Recorded evaluation configuration: {"noise\_type":"asymmetric"}. This record does not assert independent reproduction.

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