# Image Classification and Reconstruction Defense on CIFAR100 (val test)

Source: [Image Classification and Reconstruction Defense on CIFAR100 (val test)](<https://www.sota2.com/research/sota/image-classification-and-reconstruction-defense-on-cifar100-val-test>)

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

Data updated: 2026-02-26T05:30:28.752Z

Evaluation of Image Classification performance on the CIFAR-100 dataset (val/test split), specifically assessing defense mechanisms via reconstruction quality metrics (MSE).

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 and Reconstruction Defense |
| Dataset (including any stated split) | CIFAR100 (val test) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 16 of 16 |
| 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](<https://www.sota2.com/research/sota/image-classification-and-reconstruction-defense-on-cifar100-val-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [MSE (Dec., Train)](<https://www.sota2.com/research/sota/image-classification-and-reconstruction-defense-on-cifar100-val-test.md?metric=MSE%20(Dec.%2C%20Train)>): Lower is better; unit: SCALAR.

- [MSE (Dec., Infer)](<https://www.sota2.com/research/sota/image-classification-and-reconstruction-defense-on-cifar100-val-test.md?metric=MSE%20(Dec.%2C%20Infer)>): Lower is better; unit: SCALAR.

- [MSE (GAN, Train)](<https://www.sota2.com/research/sota/image-classification-and-reconstruction-defense-on-cifar100-val-test.md?metric=MSE%20(GAN%2C%20Train)>): Lower is better; unit: SCALAR.

- [MSE (GAN, Infer)](<https://www.sota2.com/research/sota/image-classification-and-reconstruction-defense-on-cifar100-val-test.md?metric=MSE%20(GAN%2C%20Infer)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Bottleneck | 68.43 | Not recorded | 68fa77c91e209131cea7556d |
| Bottleneck+CEM | 68.42 | Not recorded | 68fa77c91e209131cea7556d |
| ResSFL+CEM | 66.96 | Not recorded | 68fa77c91e209131cea7556d |
| ResSFL | 66.94 | Not recorded | 68fa77c91e209131cea7556d |
| DistCorr+CEM | 66.27 | Not recorded | 68fa77c91e209131cea7556d |
| DistCorr | 66.21 | Not recorded | 68fa77c91e209131cea7556d |
| Dropout+CEM | 65.92 | Not recorded | 68fa77c91e209131cea7556d |
| Dropout | 65.85 | Not recorded | 68fa77c91e209131cea7556d |
| Average w/o CEM | 65.81 | Not recorded | 68fa77c91e209131cea7556d |
| Average w/ CEM | 65.75 | Not recorded | 68fa77c91e209131cea7556d |
| Noise\_Nopeek | 65.55 | Not recorded | 68fa77c91e209131cea7556d |
| Noise\_Nopeek+CEM | 65.33 | Not recorded | 68fa77c91e209131cea7556d |
| PATROL | 65.1 | Not recorded | 68fa77c91e209131cea7556d |
| PATROL+CEM | 65.07 | Not recorded | 68fa77c91e209131cea7556d |
| Noise\_ARL | 62.58 | Not recorded | 68fa77c91e209131cea7556d |
| Noise\_ARL+CEM | 62.34 | Not recorded | 68fa77c91e209131cea7556d |

### Bottleneck

Accuracy: 68.43. Result ID: 699569f43badfaffe65de636.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Bottleneck+CEM

Accuracy: 68.42. Result ID: 699569f43badfaffe65de637.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### ResSFL+CEM

Accuracy: 66.96. Result ID: 699569f43badfaffe65de63f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### ResSFL

Accuracy: 66.94. Result ID: 699569f43badfaffe65de63e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### DistCorr+CEM

Accuracy: 66.27. Result ID: 699569f43badfaffe65de639.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### DistCorr

Accuracy: 66.21. Result ID: 699569f43badfaffe65de638.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Dropout+CEM

Accuracy: 65.92. Result ID: 699569f43badfaffe65de63b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Dropout

Accuracy: 65.85. Result ID: 699569f43badfaffe65de63a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Average w/o CEM

Accuracy: 65.81. Result ID: 699569f43badfaffe65de644.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Average w/ CEM

Accuracy: 65.75. Result ID: 699569f43badfaffe65de645.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Noise\_Nopeek

Accuracy: 65.55. Result ID: 699569f43badfaffe65de640.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Noise\_Nopeek+CEM

Accuracy: 65.33. Result ID: 699569f43badfaffe65de641.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### PATROL

Accuracy: 65.1. Result ID: 699569f43badfaffe65de63c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### PATROL+CEM

Accuracy: 65.07. Result ID: 699569f43badfaffe65de63d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Noise\_ARL

Accuracy: 62.58. Result ID: 699569f43badfaffe65de642.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

### Noise\_ARL+CEM

Accuracy: 62.34. Result ID: 699569f43badfaffe65de643.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c91e209131cea7556d.md>)

[Implementation](<https://github.com/xiasong0501/CEM>)

Recorded evaluation configuration: {"Backbone":"VGG11"}. This record does not assert independent reproduction.

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