# Camera Trajectory Generation on mip-NeRF 360, OMMO, Hypersim, and MannequinChallenge

Source: [Camera Trajectory Generation on mip-NeRF 360, OMMO, Hypersim, and MannequinChallenge](<https://www.sota2.com/research/sota/camera-trajectory-generation-on-mip-nerf-360-ommo-hypersim-and-mannequinchallenge>)

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

Data updated: 2026-02-26T06:07:40.296Z

Evaluation of camera pose accuracy (translation and rotation error) generated for novel view synthesis tasks across the combined set of Mip-NeRF 360, OMMO, Hypersim, and MannequinChallenge datasets.

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 Trajectory Generation |
| Dataset (including any stated split) | mip-NeRF 360, OMMO, Hypersim, and MannequinChallenge |
| Metric | Translation MSE |
| Unit | SCALAR |
| Direction | Lower 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

- [Translation MSE](<https://www.sota2.com/research/sota/camera-trajectory-generation-on-mip-nerf-360-ommo-hypersim-and-mannequinchallenge.md?metric=Translation%20MSE>): Lower is better; unit: SCALAR.

- [Rotation MSE](<https://www.sota2.com/research/sota/camera-trajectory-generation-on-mip-nerf-360-ommo-hypersim-and-mannequinchallenge.md?metric=Rotation%20MSE>): Lower is better; unit: SCALAR.

- [Visual Quality Score](<https://www.sota2.com/research/sota/camera-trajectory-generation-on-mip-nerf-360-ommo-hypersim-and-mannequinchallenge.md?metric=Visual%20Quality%20Score>): Higher is better; unit: SCALAR.

- [Alignment Score](<https://www.sota2.com/research/sota/camera-trajectory-generation-on-mip-nerf-360-ommo-hypersim-and-mannequinchallenge.md?metric=Alignment%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Translation MSE | Result date | Report paper |
| --- | --- | --- | --- |
| ChatCam | 5.3 | Not recorded | 68fa74fd1e209131cea651ba |
| ChatCam | 6.4 | Not recorded | 68fa74fd1e209131cea651ba |
| ChatCam | 7.3 | Not recorded | 68fa74fd1e209131cea651ba |
| ChatCam | 16.2 | Not recorded | 68fa74fd1e209131cea651ba |
| LERF | 17.7 | Not recorded | 68fa74fd1e209131cea651ba |
| SA3D | 19.5 | Not recorded | 68fa74fd1e209131cea651ba |

### ChatCam

Translation MSE: 5.3. Result ID: 69936d903a6c31484915dea0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

Recorded evaluation configuration: {"LLM Agent":"GPT-4","Anchor Determination":true}. This record does not assert independent reproduction.

### ChatCam

Translation MSE: 6.4. Result ID: 69936d903a6c31484915de9d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

Recorded evaluation configuration: {"LLM Agent":"LLAMA-2","Anchor Determination":true}. This record does not assert independent reproduction.

### ChatCam

Translation MSE: 7.3. Result ID: 69936d903a6c31484915de9e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

Recorded evaluation configuration: {"LLM Agent":"GPT-3.5","Anchor Determination":true}. This record does not assert independent reproduction.

### ChatCam

Translation MSE: 16.2. Result ID: 69936d903a6c31484915de9f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

Recorded evaluation configuration: {"LLM Agent":"GPT-4","Anchor Determination":false}. This record does not assert independent reproduction.

### LERF

Translation MSE: 17.7. Result ID: 69936d903a6c31484915de9c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ec61e209131cea41f85.md>)

Recorded evaluation configuration: {"LLM Agent":"GPT-4"}. This record does not assert independent reproduction.

### SA3D

Translation MSE: 19.5. Result ID: 69936d903a6c31484915de9b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74fd1e209131cea651ba.md>)

Recorded evaluation configuration: {"LLM Agent":"GPT-4"}. This record does not assert independent reproduction.

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