# Trajectory Prediction on Argoverse 2 (Pedestrian-HDV)

Source: [Trajectory Prediction on Argoverse 2 (Pedestrian-HDV)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv>)

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

Data updated: 2026-05-28T10:13:07.947Z

Evaluation of trajectory prediction models on the Pedestrian-HDV subset of the Argoverse 2 dataset, assessed using metrics covering displacement errors (ADE, FDE) and errors in predicted velocity and acceleration components (RMSE).

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) | Argoverse 2 (Pedestrian-HDV) |
| Metric | RMSE Velocity Longitudinal (m/s) |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 10 of 10 |
| 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

- [RMSE Velocity Longitudinal (m/s)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv.md?metric=RMSE%20Velocity%20Longitudinal%20(m%2Fs)>): Lower is better; unit: SCALAR.

- [RMSE Acceleration Longitudinal (m/s^2)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv.md?metric=RMSE%20Acceleration%20Longitudinal%20(m%2Fs%5E2)>): Lower is better; unit: SCALAR.

- [RMSE Velocity Lateral (m/s)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv.md?metric=RMSE%20Velocity%20Lateral%20(m%2Fs)>): Lower is better; unit: SCALAR.

- [RMSE Acceleration Lateral (m/s^2)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv.md?metric=RMSE%20Acceleration%20Lateral%20(m%2Fs%5E2)>): Lower is better; unit: SCALAR.

- [ADE (m)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv.md?metric=ADE%20(m)>): Lower is better; unit: SCALAR.

- [FDE (m)](<https://www.sota2.com/research/sota/trajectory-prediction-on-argoverse-2-pedestrian-hdv.md?metric=FDE%20(m)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | RMSE Velocity Longitudinal (m/s) | Result date | Report paper |
| --- | --- | --- | --- |
| SMamba-DDPG | 0.63 | Not recorded | 6a17cc50dbd7150506f815ad |
| Mamba-DDPG | 0.68 | Not recorded | 6a17cc50dbd7150506f815ad |
| Transformer-DDPG | 0.81 | Not recorded | 6a17cc50dbd7150506f815ad |
| LSTM-DDPG | 1.16 | Not recorded | 6a17cc50dbd7150506f815ad |
| DDPG | 1.65 | Not recorded | 6a17cc50dbd7150506f815ad |
| DQN | 1.78 | Not recorded | 6a17cc50dbd7150506f815ad |
| Transformer | 2.01 | Not recorded | 6a17cc50dbd7150506f815ad |
| DPG | 2.34 | Not recorded | 6a17cc50dbd7150506f815ad |
| LSTM | 3.19 | Not recorded | 6a17cc50dbd7150506f815ad |
| NN | 3.57 | Not recorded | 6a17cc50dbd7150506f815ad |

### SMamba-DDPG

RMSE Velocity Longitudinal (m/s): 0.63. Result ID: 6a18152efbc373592c126f26.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Hybrid Models"}. This record does not assert independent reproduction.

### Mamba-DDPG

RMSE Velocity Longitudinal (m/s): 0.68. Result ID: 6a18152efbc373592c126f27.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a871e209131cea83b8c.md>)

Recorded evaluation configuration: {"Category":"Hybrid Models"}. This record does not assert independent reproduction.

### Transformer-DDPG

RMSE Velocity Longitudinal (m/s): 0.81. Result ID: 6a18152efbc373592c126f28.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Hybrid Models"}. This record does not assert independent reproduction.

### LSTM-DDPG

RMSE Velocity Longitudinal (m/s): 1.16. Result ID: 6a18152efbc373592c126f29.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Hybrid Models"}. This record does not assert independent reproduction.

### DDPG

RMSE Velocity Longitudinal (m/s): 1.65. Result ID: 6a18152efbc373592c126f2a.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65fd1e209131cea0dc8e.md>)

Recorded evaluation configuration: {"Category":"Reinforcement Learning Baselines"}. This record does not assert independent reproduction.

### DQN

RMSE Velocity Longitudinal (m/s): 1.78. Result ID: 6a18152efbc373592c126f2b.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Reinforcement Learning Baselines"}. This record does not assert independent reproduction.

### Transformer

RMSE Velocity Longitudinal (m/s): 2.01. Result ID: 6a18152efbc373592c126f2d.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Supervised Learning Baselines"}. This record does not assert independent reproduction.

### DPG

RMSE Velocity Longitudinal (m/s): 2.34. Result ID: 6a18152efbc373592c126f2c.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Reinforcement Learning Baselines"}. This record does not assert independent reproduction.

### LSTM

RMSE Velocity Longitudinal (m/s): 3.19. Result ID: 6a18152efbc373592c126f2e.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Supervised Learning Baselines"}. This record does not assert independent reproduction.

### NN

RMSE Velocity Longitudinal (m/s): 3.57. Result ID: 6a18152efbc373592c126f2f.

[Reporting paper](<https://www.sota2.com/research/paper/6a17cc50dbd7150506f815ad.md>)

Recorded evaluation configuration: {"Category":"Supervised Learning Baselines"}. This record does not assert independent reproduction.

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