Improving Out-of-Distribution Generalization of Trajectory Prediction for Autonomous Driving via Polynomial Representations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909466138509312 |
|---|---|
| author | Yao, Yue Yan, Shengchao Goehring, Daniel Burgard, Wolfram Reichardt, Joerg |
| author_facet | Yao, Yue Yan, Shengchao Goehring, Daniel Burgard, Wolfram Reichardt, Joerg |
| contents | Robustness against Out-of-Distribution (OoD) samples is a key performance indicator of a trajectory prediction model. However, the development and ranking of state-of-the-art (SotA) models are driven by their In-Distribution (ID) performance on individual competition datasets. We present an OoD testing protocol that homogenizes datasets and prediction tasks across two large-scale motion datasets. We introduce a novel prediction algorithm based on polynomial representations for agent trajectory and road geometry on both the input and output sides of the model. With a much smaller model size, training effort, and inference time, we reach near SotA performance for ID testing and significantly improve robustness in OoD testing. Within our OoD testing protocol, we further study two augmentation strategies of SotA models and their effects on model generalization. Highlighting the contrast between ID and OoD performance, we suggest adding OoD testing to the evaluation criteria of trajectory prediction models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_13431 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Improving Out-of-Distribution Generalization of Trajectory Prediction for Autonomous Driving via Polynomial Representations Yao, Yue Yan, Shengchao Goehring, Daniel Burgard, Wolfram Reichardt, Joerg Machine Learning Artificial Intelligence Robustness against Out-of-Distribution (OoD) samples is a key performance indicator of a trajectory prediction model. However, the development and ranking of state-of-the-art (SotA) models are driven by their In-Distribution (ID) performance on individual competition datasets. We present an OoD testing protocol that homogenizes datasets and prediction tasks across two large-scale motion datasets. We introduce a novel prediction algorithm based on polynomial representations for agent trajectory and road geometry on both the input and output sides of the model. With a much smaller model size, training effort, and inference time, we reach near SotA performance for ID testing and significantly improve robustness in OoD testing. Within our OoD testing protocol, we further study two augmentation strategies of SotA models and their effects on model generalization. Highlighting the contrast between ID and OoD performance, we suggest adding OoD testing to the evaluation criteria of trajectory prediction models. |
| title | Improving Out-of-Distribution Generalization of Trajectory Prediction for Autonomous Driving via Polynomial Representations |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2407.13431 |