# Watermark Extraction on MS-COCO 2017 (test)

Source: [Watermark Extraction on MS-COCO 2017 (test)](<https://www.sota2.com/research/sota/watermark-extraction-on-ms-coco-2017-test>)

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

Data updated: 2026-02-26T05:39:18.204Z

Evaluation of Watermark Extraction performance on the test split of the MS-COCO 2017 dataset, reporting metrics such as Clean Accuracy, Adversarial Accuracy, and Payload Rate.

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 | Watermark Extraction |
| Dataset (including any stated split) | MS-COCO 2017 (test) |
| Metric | Clean Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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/watermark-extraction-on-ms-coco-2017-test.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [Adversarial Accuracy](<https://www.sota2.com/research/sota/watermark-extraction-on-ms-coco-2017-test.md?metric=Adversarial%20Accuracy>): Higher is better; unit: PERCENT.

- [Payload Rate](<https://www.sota2.com/research/sota/watermark-extraction-on-ms-coco-2017-test.md?metric=Payload%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| GS | 100 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| OSI | 100 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| OSI | 99.97 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| GS | 99.92 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| OSI | 95.9 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| GS | 94.73 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| OSI | 88.15 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |
| GS | 86.52 | Not recorded | 698c0d4cc97a0b3e3b5e1304 |

### GS

Clean Accuracy: 100. Result ID: 69971f7e5698665ec46da97c.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73131e209131cea5a4a6.md>)

Recorded evaluation configuration: {"f\_hw":8}. This record does not assert independent reproduction.

### OSI

Clean Accuracy: 100. Result ID: 69971f7e5698665ec46da980.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

Recorded evaluation configuration: {"f\_hw":8}. This record does not assert independent reproduction.

### OSI

Clean Accuracy: 99.97. Result ID: 69971f7e5698665ec46da97f.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

Recorded evaluation configuration: {"f\_hw":4}. This record does not assert independent reproduction.

### GS

Clean Accuracy: 99.92. Result ID: 69971f7e5698665ec46da97b.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73131e209131cea5a4a6.md>)

Recorded evaluation configuration: {"f\_hw":4}. This record does not assert independent reproduction.

### OSI

Clean Accuracy: 95.9. Result ID: 69971f7e5698665ec46da97e.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

Recorded evaluation configuration: {"f\_hw":2}. This record does not assert independent reproduction.

### GS

Clean Accuracy: 94.73. Result ID: 69971f7e5698665ec46da97a.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73131e209131cea5a4a6.md>)

Recorded evaluation configuration: {"f\_hw":2}. This record does not assert independent reproduction.

### OSI

Clean Accuracy: 88.15. Result ID: 69971f7e5698665ec46da97d.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

Recorded evaluation configuration: {"f\_hw":1}. This record does not assert independent reproduction.

### GS

Clean Accuracy: 86.52. Result ID: 69971f7e5698665ec46da979.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d4cc97a0b3e3b5e1304.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73131e209131cea5a4a6.md>)

Recorded evaluation configuration: {"f\_hw":1}. This record does not assert independent reproduction.

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