# Trajectory Prediction on RELLIS-3D

Source: [Trajectory Prediction on RELLIS-3D](<https://www.sota2.com/research/sota/trajectory-prediction-on-rellis-3d>)

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

Data updated: 2026-02-26T05:48:01.759Z

Evaluation of trajectory prediction models on the RELLIS-3D dataset, typically assessed using L2 error (in meters) over various prediction horizons (1s, 2s, 3s) and overall failure 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 | Trajectory Prediction |
| Dataset (including any stated split) | RELLIS-3D |
| Metric | L2 Error (1s) |
| Unit | SCALAR |
| Direction | Lower 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

- [L2 Error (1s)](<https://www.sota2.com/research/sota/trajectory-prediction-on-rellis-3d.md?metric=L2%20Error%20(1s)>): Lower is better; unit: SCALAR.

- [L2 Error (2s)](<https://www.sota2.com/research/sota/trajectory-prediction-on-rellis-3d.md?metric=L2%20Error%20(2s)>): Lower is better; unit: SCALAR.

- [L2 Error (3s)](<https://www.sota2.com/research/sota/trajectory-prediction-on-rellis-3d.md?metric=L2%20Error%20(3s)>): Lower is better; unit: SCALAR.

- [L2 Error (Avg)](<https://www.sota2.com/research/sota/trajectory-prediction-on-rellis-3d.md?metric=L2%20Error%20(Avg)>): Lower is better; unit: SCALAR.

- [Failure Rate](<https://www.sota2.com/research/sota/trajectory-prediction-on-rellis-3d.md?metric=Failure%20Rate>): Lower is better; unit: PERCENT.

## Recorded results

| Method | L2 Error (1s) | Result date | Report paper |
| --- | --- | --- | --- |
| OFF-EMMA | 0.88 | Not recorded | 695f39e453fdd5c361743d3b |
| LightEMMA | 0.93 | Not recorded | 695f39e453fdd5c361743d3b |
| OpenEMMA | 0.95 | Not recorded | 695f39e453fdd5c361743d3b |
| OFF-EMMA | 1.17 | Not recorded | 695f39e453fdd5c361743d3b |
| OpenEMMA | 1.23 | Not recorded | 695f39e453fdd5c361743d3b |
| LightEMMA | 1.26 | Not recorded | 695f39e453fdd5c361743d3b |
| Zero-shot | 1.39 | Not recorded | 695f39e453fdd5c361743d3b |
| Zero-shot | 1.64 | Not recorded | 695f39e453fdd5c361743d3b |

### OFF-EMMA

L2 Error (1s): 0.88. Result ID: 6999dd035698665ec45c4731.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-VL-7B-Instruct"}. This record does not assert independent reproduction.

### LightEMMA

L2 Error (1s): 0.93. Result ID: 6999dd035698665ec45c472f.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b461e209131cea87731.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-VL-7B-Instruct"}. This record does not assert independent reproduction.

### OpenEMMA

L2 Error (1s): 0.95. Result ID: 6999dd035698665ec45c472d.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77071e209131cea709b4.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-VL-7B-Instruct"}. This record does not assert independent reproduction.

### OFF-EMMA

L2 Error (1s): 1.17. Result ID: 6999dd035698665ec45c4730.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

Recorded evaluation configuration: {"Model":"Llama-3.2-11B-Vision-Instruct"}. This record does not assert independent reproduction.

### OpenEMMA

L2 Error (1s): 1.23. Result ID: 6999dd035698665ec45c472c.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77071e209131cea709b4.md>)

Recorded evaluation configuration: {"Model":"Llama-3.2-11B-Vision-Instruct"}. This record does not assert independent reproduction.

### LightEMMA

L2 Error (1s): 1.26. Result ID: 6999dd035698665ec45c472e.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b461e209131cea87731.md>)

Recorded evaluation configuration: {"Model":"Llama-3.2-11B-Vision-Instruct"}. This record does not assert independent reproduction.

### Zero-shot

L2 Error (1s): 1.39. Result ID: 6999dd035698665ec45c472b.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-VL-7B-Instruct"}. This record does not assert independent reproduction.

### Zero-shot

L2 Error (1s): 1.64. Result ID: 6999dd035698665ec45c472a.

[Reporting paper](<https://www.sota2.com/research/paper/695f39e453fdd5c361743d3b.md>)

Recorded evaluation configuration: {"Model":"Llama-3.2-11B-Vision-Instruct"}. This record does not assert independent reproduction.

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