# Camera-to-gravity Rotation Estimation on ETH3D

Source: [Camera-to-gravity Rotation Estimation on ETH3D](<https://www.sota2.com/research/sota/camera-to-gravity-rotation-estimation-on-eth3d>)

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

Data updated: 2026-05-27T10:35:45.684Z

Evaluation of camera orientation estimation relative to the gravity vector on the ETH3D dataset, reporting rotational error statistics (Rerr) and rotational accuracy (Racc) at various angular thresholds.

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 | Camera-to-gravity rotation estimation |
| Dataset (including any stated split) | ETH3D |
| Metric | Mean Rotational Error |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 9 of 9 |
| 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

- [Mean Rotational Error](<https://www.sota2.com/research/sota/camera-to-gravity-rotation-estimation-on-eth3d.md?metric=Mean%20Rotational%20Error>): Lower is better; unit: SCALAR.

- [Median Rotational Error](<https://www.sota2.com/research/sota/camera-to-gravity-rotation-estimation-on-eth3d.md?metric=Median%20Rotational%20Error>): Lower is better; unit: SCALAR.

- [Accuracy @1°](<https://www.sota2.com/research/sota/camera-to-gravity-rotation-estimation-on-eth3d.md?metric=Accuracy%20%401%C2%B0>): Higher is better; unit: PERCENT.

- [Accuracy @2°](<https://www.sota2.com/research/sota/camera-to-gravity-rotation-estimation-on-eth3d.md?metric=Accuracy%20%402%C2%B0>): Higher is better; unit: PERCENT.

- [Accuracy @5°](<https://www.sota2.com/research/sota/camera-to-gravity-rotation-estimation-on-eth3d.md?metric=Accuracy%20%405%C2%B0>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Mean Rotational Error | Result date | Report paper |
| --- | --- | --- | --- |
| Local Head | 1.05 | Not recorded | 6a167adbdbd7150506f8033c |
| Local Head | 1.11 | Not recorded | 6a167adbdbd7150506f8033c |
| Local Head | 1.62 | Not recorded | 6a167adbdbd7150506f8033c |
| Procrustes | 1.85 | Not recorded | 6a167adbdbd7150506f8033c |
| Procrustes | 1.95 | Not recorded | 6a167adbdbd7150506f8033c |
| Procrustes | 1.96 | Not recorded | 6a167adbdbd7150506f8033c |
| GeoCalib | 2.21 | Not recorded | 6a167adbdbd7150506f8033c |
| GeoCalib | 2.22 | Not recorded | 6a167adbdbd7150506f8033c |
| GeoCalib | 2.24 | Not recorded | 6a167adbdbd7150506f8033c |

### Local Head

Mean Rotational Error: 1.05. Result ID: 6a16c8eafbc373592c10a322.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"8"}. This record does not assert independent reproduction.

### Local Head

Mean Rotational Error: 1.11. Result ID: 6a16c8eafbc373592c10a31f.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"4"}. This record does not assert independent reproduction.

### Local Head

Mean Rotational Error: 1.62. Result ID: 6a16c8eafbc373592c10a31c.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"1"}. This record does not assert independent reproduction.

### Procrustes

Mean Rotational Error: 1.85. Result ID: 6a16c8eafbc373592c10a321.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"8"}. This record does not assert independent reproduction.

### Procrustes

Mean Rotational Error: 1.95. Result ID: 6a16c8eafbc373592c10a31e.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"4"}. This record does not assert independent reproduction.

### Procrustes

Mean Rotational Error: 1.96. Result ID: 6a16c8eafbc373592c10a31b.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"1"}. This record does not assert independent reproduction.

### GeoCalib

Mean Rotational Error: 2.21. Result ID: 6a16c8eafbc373592c10a320.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"8"}. This record does not assert independent reproduction.

### GeoCalib

Mean Rotational Error: 2.22. Result ID: 6a16c8eafbc373592c10a31d.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"4"}. This record does not assert independent reproduction.

### GeoCalib

Mean Rotational Error: 2.24. Result ID: 6a16c8eafbc373592c10a31a.

[Reporting paper](<https://www.sota2.com/research/paper/6a167adbdbd7150506f8033c.md>)

Recorded evaluation configuration: {"Number of input views":"1"}. This record does not assert independent reproduction.

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