# Image-level Anomaly Detection on R2T (test)

Source: [Image-level Anomaly Detection on R2T (test)](<https://www.sota2.com/research/sota/image-level-anomaly-detection-on-r2t-test>)

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

Data updated: 2026-05-26T10:59:09.575Z

Evaluation of Image-level Anomaly Detection performance on the test split of the R2T dataset, reporting I-AUROC, I-AP, and I-F1 scores.

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-level Anomaly Detection |
| Dataset (including any stated split) | R2T (test) |
| Metric | I-AUROC |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 13 of 13 |
| 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

- [I-AUROC](<https://www.sota2.com/research/sota/image-level-anomaly-detection-on-r2t-test.md?metric=I-AUROC>): Higher is better; unit: PERCENT.

- [I-AP](<https://www.sota2.com/research/sota/image-level-anomaly-detection-on-r2t-test.md?metric=I-AP>): Higher is better; unit: PERCENT.

- [I-F1](<https://www.sota2.com/research/sota/image-level-anomaly-detection-on-r2t-test.md?metric=I-F1>): Higher is better; unit: PERCENT.

## Recorded results

| Method | I-AUROC | Result date | Report paper |
| --- | --- | --- | --- |
| AVATAR | 71.15 | Not recorded | 6a152967dbd7150506f7e35c |
| AnomalyCLIP | 69.72 | Not recorded | 6a152967dbd7150506f7e35c |
| WinCLIP | 67.31 | Not recorded | 6a152967dbd7150506f7e35c |
| Bayes-PFL | 64.45 | Not recorded | 6a152967dbd7150506f7e35c |
| Gradient Residual | 62.25 | Not recorded | 6a152967dbd7150506f7e35c |
| APRIL-GAN | 61.07 | Not recorded | 6a152967dbd7150506f7e35c |
| SSIM Residual | 59.64 | Not recorded | 6a152967dbd7150506f7e35c |
| UniAD | 57.29 | Not recorded | 6a152967dbd7150506f7e35c |
| Norm. RGB Residual | 47.88 | Not recorded | 6a152967dbd7150506f7e35c |
| AA-CLIP | 39.08 | Not recorded | 6a152967dbd7150506f7e35c |
| AdaCLIP | 38.26 | Not recorded | 6a152967dbd7150506f7e35c |
| Dinomaly | 37.85 | Not recorded | 6a152967dbd7150506f7e35c |
| INP-Former | 31.76 | Not recorded | 6a152967dbd7150506f7e35c |

### AVATAR

I-AUROC: 71.15. Result ID: 6a157c54fbc373592c0dbb1b.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### AnomalyCLIP

I-AUROC: 69.72. Result ID: 6a157c54fbc373592c0dbb18.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78451e209131cea77851.md>)

[Implementation](<https://github.com/zqhang/AnomalyCLIP>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### WinCLIP

I-AUROC: 67.31. Result ID: 6a157c54fbc373592c0dbb15.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6eea1e209131cea42e2d.md>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### Bayes-PFL

I-AUROC: 64.45. Result ID: 6a157c54fbc373592c0dbb1a.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79151e209131cea7c094.md>)

[Implementation](<https://github.com/xiaozhen228/Bayes-PFL>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### Gradient Residual

I-AUROC: 62.25. Result ID: 6a157c54fbc373592c0dbb10.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

Recorded evaluation configuration: {"Type":"T"}. This record does not assert independent reproduction.

### APRIL-GAN

I-AUROC: 61.07. Result ID: 6a157c54fbc373592c0dbb17.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d71e209131cea4d95d.md>)

[Implementation](<https://github.com/ByChelsea/VAND-APRIL-GAN>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### SSIM Residual

I-AUROC: 59.64. Result ID: 6a157c54fbc373592c0dbb11.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

Recorded evaluation configuration: {"Type":"T"}. This record does not assert independent reproduction.

### UniAD

I-AUROC: 57.29. Result ID: 6a157c54fbc373592c0dbb12.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dba1e209131cea3b9bd.md>)

[Implementation](<https://github.com/zhiyuanyou/UniAD>)

Recorded evaluation configuration: {"Type":"U"}. This record does not assert independent reproduction.

### Norm. RGB Residual

I-AUROC: 47.88. Result ID: 6a157c54fbc373592c0dbb0f.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

Recorded evaluation configuration: {"Type":"T"}. This record does not assert independent reproduction.

### AA-CLIP

I-AUROC: 39.08. Result ID: 6a157c54fbc373592c0dbb19.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77551e209131cea7270b.md>)

[Implementation](<https://github.com/Mwxinnn/AA-CLIP>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### AdaCLIP

I-AUROC: 38.26. Result ID: 6a157c54fbc373592c0dbb16.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77061e209131cea708e2.md>)

[Implementation](<https://github.com/caoyunkang/AdaCLIP>)

Recorded evaluation configuration: {"Type":"Z"}. This record does not assert independent reproduction.

### Dinomaly

I-AUROC: 37.85. Result ID: 6a157c54fbc373592c0dbb13.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77c51e209131cea75373.md>)

[Implementation](<https://github.com/guojiajeremy/Dinomaly>)

Recorded evaluation configuration: {"Type":"U"}. This record does not assert independent reproduction.

### INP-Former

I-AUROC: 31.76. Result ID: 6a157c54fbc373592c0dbb14.

[Reporting paper](<https://www.sota2.com/research/paper/6a152967dbd7150506f7e35c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79cc1e209131cea7f688.md>)

[Implementation](<https://github.com/luow23/INP-Former>)

Recorded evaluation configuration: {"Type":"U"}. This record does not assert independent reproduction.

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