# Open-Vocabulary 3D Object Detection on SUN RGB-D (test)

Source: [Open-Vocabulary 3D Object Detection on SUN RGB-D (test)](<https://www.sota2.com/research/sota/open-vocabulary-3d-object-detection-on-sun-rgb-d-test>)

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

Data updated: 2026-02-27T09:40:34.453Z

Evaluation of Open-Vocabulary 3D Object Detection performance on the test split of the SUN RGB-D dataset, reporting metrics for novel, base, and all classes.

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 | Open-Vocabulary 3D Object Detection |
| Dataset (including any stated split) | SUN RGB-D (test) |
| Metric | AP Novel |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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 Novel](<https://www.sota2.com/research/sota/open-vocabulary-3d-object-detection-on-sun-rgb-d-test.md?metric=AP%20Novel>): Higher is better; unit: PERCENT.

- [AP Base](<https://www.sota2.com/research/sota/open-vocabulary-3d-object-detection-on-sun-rgb-d-test.md?metric=AP%20Base>): Higher is better; unit: PERCENT.

- [AP All](<https://www.sota2.com/research/sota/open-vocabulary-3d-object-detection-on-sun-rgb-d-test.md?metric=AP%20All>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AP Novel | Result date | Report paper |
| --- | --- | --- | --- |
| OneDet3D | 12.59 | Not recorded | 68fa75841e209131cea68532 |
| OneDet3D | 11.01 | Not recorded | 68fa75841e209131cea68532 |
| CoDA | 6.65 | Not recorded | 68fa75841e209131cea68532 |
| 3D-CLIP | 3.02 | Not recorded | 68fa75841e209131cea68532 |
| Det-PointCLIPv2 | 0.12 | Not recorded | 68fa75841e209131cea68532 |
| Det-PointCLIP | 0.08 | Not recorded | 68fa75841e209131cea68532 |

### OneDet3D

AP Novel: 12.59. Result ID: 69a1668335dfcabb00e6d07c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

Recorded evaluation configuration: {"Training Protocol":"multi-dataset training"}. This record does not assert independent reproduction.

### OneDet3D

AP Novel: 11.01. Result ID: 69a1668335dfcabb00e6d07b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

Recorded evaluation configuration: {"Training Protocol":"single-dataset training"}. This record does not assert independent reproduction.

### CoDA

AP Novel: 6.65. Result ID: 69a1668335dfcabb00e6d07a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70ce1e209131cea4d4db.md>)

Recorded evaluation configuration: {"Training Protocol":"single-dataset training"}. This record does not assert independent reproduction.

### 3D-CLIP

AP Novel: 3.02. Result ID: 69a1668335dfcabb00e6d079.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

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

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

Recorded evaluation configuration: {"Training Protocol":"single-dataset training"}. This record does not assert independent reproduction.

### Det-PointCLIPv2

AP Novel: 0.12. Result ID: 69a1668335dfcabb00e6d078.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70701e209131cea4b002.md>)

[Implementation](<https://github.com/yangyangyang127/PointCLIP_V2>)

Recorded evaluation configuration: {"Training Protocol":"single-dataset training"}. This record does not assert independent reproduction.

### Det-PointCLIP

AP Novel: 0.08. Result ID: 69a1668335dfcabb00e6d077.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75841e209131cea68532.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b5d1e209131cea2dabe.md>)

[Implementation](<https://github.com/ZrrSkywalker/PointCLIP>)

Recorded evaluation configuration: {"Training Protocol":"single-dataset training"}. This record does not assert independent reproduction.

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