# Trajectory Prediction on TUM intersection dataset 2s horizon

Source: [Trajectory Prediction on TUM intersection dataset 2s horizon](<https://www.sota2.com/research/sota/trajectory-prediction-on-tum-intersection-dataset-2s-horizon>)

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

Data updated: 2026-03-09T07:03:53.733Z

Evaluation of trajectory prediction models on the TUM intersection dataset, typically assessed using metrics like ADE, FDE, RMSE, and NL-ADE.

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) | TUM intersection dataset 2s horizon |
| Metric | ADE |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| 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

- [ADE](<https://www.sota2.com/research/sota/trajectory-prediction-on-tum-intersection-dataset-2s-horizon.md?metric=ADE>): Lower is better; unit: SCALAR.

- [FDE](<https://www.sota2.com/research/sota/trajectory-prediction-on-tum-intersection-dataset-2s-horizon.md?metric=FDE>): Lower is better; unit: SCALAR.

- [RMSE](<https://www.sota2.com/research/sota/trajectory-prediction-on-tum-intersection-dataset-2s-horizon.md?metric=RMSE>): Lower is better; unit: SCALAR.

- [NL-ADE](<https://www.sota2.com/research/sota/trajectory-prediction-on-tum-intersection-dataset-2s-horizon.md?metric=NL-ADE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | ADE | Result date | Report paper |
| --- | --- | --- | --- |
| Baseline | 0.548 | Not recorded | 69ae53f52d603e6de3e429f1 |
| Diversity\_Loss | 0.548 | Not recorded | 69ae53f52d603e6de3e429f1 |
| Map\_Loss | 0.716 | Not recorded | 69ae53f52d603e6de3e429f1 |
| Collision\_Loss | 0.778 | Not recorded | 69ae53f52d603e6de3e429f1 |
| CV | 0.931 | Not recorded | 69ae53f52d603e6de3e429f1 |
| Twin\_All | 0.936 | Not recorded | 69ae53f52d603e6de3e429f1 |
| KF-CA | 1.001 | Not recorded | 69ae53f52d603e6de3e429f1 |

### Baseline

ADE: 0.548. Result ID: 69ae70d06dbb7075b9560698.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Extra losses":"MSE only","Description":"Standard LSTM baseline"}. This record does not assert independent reproduction.

### Diversity\_Loss

ADE: 0.548. Result ID: 69ae70d06dbb7075b956069c.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Description":"Stochastic best sample variant"}. This record does not assert independent reproduction.

### Map\_Loss

ADE: 0.716. Result ID: 69ae70d06dbb7075b9560699.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Extra losses":"L\_infra (Lane proximity)"}. This record does not assert independent reproduction.

### Collision\_Loss

ADE: 0.778. Result ID: 69ae70d06dbb7075b956069a.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Extra losses":"L\_coll (Batch repulsion)"}. This record does not assert independent reproduction.

### CV

ADE: 0.931. Result ID: 69ae70d06dbb7075b9560696.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Description":"Constant velocity","Training":"no training"}. This record does not assert independent reproduction.

### Twin\_All

ADE: 0.936. Result ID: 69ae70d06dbb7075b956069b.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Extra losses":"L\_infra + L\_coll","Description":"Both digital-twin losses"}. This record does not assert independent reproduction.

### KF-CA

ADE: 1.001. Result ID: 69ae70d06dbb7075b9560697.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53f52d603e6de3e429f1.md>)

Recorded evaluation configuration: {"Description":"Kalman-Singer","Training":"no training"}. This record does not assert independent reproduction.

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