# Image Classification on Food101 (Clean and Robust Accuracy)

Source: [Image Classification on Food101 (Clean and Robust Accuracy)](<https://www.sota2.com/research/sota/image-classification-on-food101-clean-and-robust-accuracy>)

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

Data updated: 2026-02-27T11:19:52.331Z

Evaluation of Image Classification performance on the Food101 dataset, specifically reporting Clean Accuracy and Robust Accuracy metrics.

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) | Food101 |
| Metric | Clean Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 28 of 28 |
| 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

- [Clean Accuracy](<https://www.sota2.com/research/sota/image-classification-on-food101-clean-and-robust-accuracy.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [Robust Accuracy](<https://www.sota2.com/research/sota/image-classification-on-food101-clean-and-robust-accuracy.md?metric=Robust%20Accuracy>): Higher is better; unit: SCALAR.

- [Adversarial Accuracy](<https://www.sota2.com/research/sota/image-classification-on-food101-clean-and-robust-accuracy.md?metric=Adversarial%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| CLIP | 90.8 | Not recorded | 69799806a2b28e761d87cb52 |
| R-TPT | 90.6 | Not recorded | 69799806a2b28e761d87cb52 |
| CSR | 90.6 | Not recorded | 69799806a2b28e761d87cb52 |
| TTE | 89.8 | Not recorded | 69799806a2b28e761d87cb52 |
| HD | 88.7 | Not recorded | 69799806a2b28e761d87cb52 |
| Anti-Adv | 87.7 | Not recorded | 69799806a2b28e761d87cb52 |
| LPF | 86.8 | Not recorded | 69799806a2b28e761d87cb52 |
| CLIP | 82.93 | Not recorded | 69705d96e25ad2fa9d1ed1a8 |
| R-TPT | 82.1 | Not recorded | 69799806a2b28e761d87cb52 |
| CLIP | 80.4 | Not recorded | 69799806a2b28e761d87cb52 |
| TTE | 79.9 | Not recorded | 69799806a2b28e761d87cb52 |
| HD | 79.7 | Not recorded | 69799806a2b28e761d87cb52 |
| CSR | 79.4 | Not recorded | 69799806a2b28e761d87cb52 |
| Anti-Adv | 78.5 | Not recorded | 69799806a2b28e761d87cb52 |
| HPT-GPD | 75.05 | Not recorded | 69705d96e25ad2fa9d1ed1a8 |
| TTC | 72.7 | Not recorded | 69799806a2b28e761d87cb52 |
| LPF | 71 | Not recorded | 69799806a2b28e761d87cb52 |
| PMG | 70.52 | Not recorded | 69705d96e25ad2fa9d1ed1a8 |
| FT-Standard | 70.07 | Not recorded | 69705d96e25ad2fa9d1ed1a8 |
| TTC | 68.3 | Not recorded | 69799806a2b28e761d87cb52 |
| FT-TeCoA | 52.81 | Not recorded | 69705d96e25ad2fa9d1ed1a8 |
| FARE | 49.4 | Not recorded | 69799806a2b28e761d87cb52 |
| FARE | 38.3 | Not recorded | 69799806a2b28e761d87cb52 |
| TeCoA | 36.2 | Not recorded | 69799806a2b28e761d87cb52 |
| TeCoA | 24.1 | Not recorded | 69799806a2b28e761d87cb52 |

### CLIP

Clean Accuracy: 90.8. Result ID: 69a17d1735dfcabb00e762bc.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Original"}. This record does not assert independent reproduction.

### R-TPT

Clean Accuracy: 90.6. Result ID: 69a17d1735dfcabb00e762bf.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7add1e209131cea8590d.md>)

[Implementation](<https://github.com/TomSheng21/R-TPT>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### CSR

Clean Accuracy: 90.6. Result ID: 69a17d1735dfcabb00e762c5.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### TTE

Clean Accuracy: 89.8. Result ID: 69a17d1735dfcabb00e762c3.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a711e209131cea2817e.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### HD

Clean Accuracy: 88.7. Result ID: 69a17d1735dfcabb00e762c1.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a1d1e209131cea26045.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### Anti-Adv

Clean Accuracy: 87.7. Result ID: 69a17d1735dfcabb00e762c2.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e9af.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### LPF

Clean Accuracy: 86.8. Result ID: 69a17d1735dfcabb00e762c0.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### CLIP

Clean Accuracy: 82.93. Result ID: 69a17dcc35dfcabb00e7658d.

[Reporting paper](<https://www.sota2.com/research/paper/69705d96e25ad2fa9d1ed1a8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Backbone":"ViT-B/32","Proxy CLIP":"ViT-B/16","Attack Protocol":"PGD-10","Training time per epoch (s)":"0"}. This record does not assert independent reproduction.

### R-TPT

Clean Accuracy: 82.1. Result ID: 69a17d1735dfcabb00e76162.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7add1e209131cea8590d.md>)

[Implementation](<https://github.com/TomSheng21/R-TPT>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### CLIP

Clean Accuracy: 80.4. Result ID: 69a17d1735dfcabb00e7615f.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Original"}. This record does not assert independent reproduction.

### TTE

Clean Accuracy: 79.9. Result ID: 69a17d1735dfcabb00e76166.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a711e209131cea2817e.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### HD

Clean Accuracy: 79.7. Result ID: 69a17d1735dfcabb00e76164.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a1d1e209131cea26045.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### CSR

Clean Accuracy: 79.4. Result ID: 69a17d1735dfcabb00e76168.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### Anti-Adv

Clean Accuracy: 78.5. Result ID: 69a17d1735dfcabb00e76165.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e9af.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### HPT-GPD

Clean Accuracy: 75.05. Result ID: 69a17dcc35dfcabb00e76591.

[Reporting paper](<https://www.sota2.com/research/paper/69705d96e25ad2fa9d1ed1a8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69705d96e25ad2fa9d1ed1a8.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B/32","Proxy CLIP":"ViT-B/16","Attack Protocol":"PGD-10","Training time per epoch (s)":"1486"}. This record does not assert independent reproduction.

### TTC

Clean Accuracy: 72.7. Result ID: 69a17d1735dfcabb00e76167.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa774a1e209131cea724cc.md>)

[Implementation](<https://www.sota2.com/here>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### LPF

Clean Accuracy: 71. Result ID: 69a17d1735dfcabb00e76163.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Test-Time Defense"}. This record does not assert independent reproduction.

### PMG

Clean Accuracy: 70.52. Result ID: 69a17dcc35dfcabb00e76590.

[Reporting paper](<https://www.sota2.com/research/paper/69705d96e25ad2fa9d1ed1a8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72df1e209131cea5912e.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B/32","Proxy CLIP":"ViT-B/16","Attack Protocol":"PGD-10","Training time per epoch (s)":"1817"}. This record does not assert independent reproduction.

### FT-Standard

Clean Accuracy: 70.07. Result ID: 69a17dcc35dfcabb00e7658e.

[Reporting paper](<https://www.sota2.com/research/paper/69705d96e25ad2fa9d1ed1a8.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B/32","Proxy CLIP":"ViT-B/16","Attack Protocol":"PGD-10","Training time per epoch (s)":"406"}. This record does not assert independent reproduction.

### TTC

Clean Accuracy: 68.3. Result ID: 69a17d1735dfcabb00e762c4.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa774a1e209131cea724cc.md>)

[Implementation](<https://www.sota2.com/here>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Test-Time Defense"}. This record does not assert independent reproduction.

### FT-TeCoA

Clean Accuracy: 52.81. Result ID: 69a17dcc35dfcabb00e7658f.

[Reporting paper](<https://www.sota2.com/research/paper/69705d96e25ad2fa9d1ed1a8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f271e209131cea44177.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B/32","Proxy CLIP":"ViT-B/16","Attack Protocol":"PGD-10","Training time per epoch (s)":"1112"}. This record does not assert independent reproduction.

### FARE

Clean Accuracy: 49.4. Result ID: 69a17d1735dfcabb00e762be.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa738c1e209131cea5d24a.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Adversarial Fine-tuning"}. This record does not assert independent reproduction.

### FARE

Clean Accuracy: 38.3. Result ID: 69a17d1735dfcabb00e76161.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa738c1e209131cea5d24a.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Adversarial Fine-tuning"}. This record does not assert independent reproduction.

### TeCoA

Clean Accuracy: 36.2. Result ID: 69a17d1735dfcabb00e762bd.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f271e209131cea44177.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-L/14","Method category":"Adversarial Fine-tuning"}. This record does not assert independent reproduction.

### TeCoA

Clean Accuracy: 24.1. Result ID: 69a17d1735dfcabb00e76160.

[Reporting paper](<https://www.sota2.com/research/paper/69799806a2b28e761d87cb52.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f271e209131cea44177.md>)

Recorded evaluation configuration: {"Backbone":"CLIP-B/32","Category":"Adversarial Fine-tuning"}. This record does not assert independent reproduction.

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