# Region Captioning on Visual Genome cleansed (test)

Source: [Region Captioning on Visual Genome cleansed (test)](<https://www.sota2.com/research/sota/region-captioning-on-visual-genome-cleansed-test>)

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

Data updated: 2026-02-26T05:22:37.250Z

Evaluation of Region Captioning performance on the test split of the cleansed Visual Genome dataset, measured using Average Precision (AP).

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 | Region Captioning |
| Dataset (including any stated split) | Visual Genome cleansed (test) |
| Metric | AP |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [AP](<https://www.sota2.com/research/sota/region-captioning-on-visual-genome-cleansed-test.md?metric=AP>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AP | Result date | Report paper |
| --- | --- | --- | --- |
| MSDN | 5.39 | Not recorded | 68fa63a51e209131cea000d9 |
| Baseline | 4.41 | Not recorded | 68fa63a51e209131cea000d9 |
| Baseline-3-bran. | 4.28 | Not recorded | 68fa63a51e209131cea000d9 |

### MSDN

AP: 5.39. Result ID: 698bfe94f729bc7f68e765d7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63a51e209131cea000d9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63a51e209131cea000d9.md>)

[Implementation](<https://github.com/yikang-li/MSDN>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Baseline

AP: 4.41. Result ID: 698bfe94f729bc7f68e765d5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63a51e209131cea000d9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62311e209131ce9f8239.md>)

[Implementation](<https://github.com/qassemoquab/stnbhwd>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Baseline-3-bran.

AP: 4.28. Result ID: 698bfe94f729bc7f68e765d6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63a51e209131cea000d9.md>)

Recorded evaluation configuration: {"branches":"3 separate branches","message passing":"false"}. This record does not assert independent reproduction.

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