MetaTra: Meta-Learning for Generalized Trajectory Prediction in Unseen Domain
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914677559132160 |
|---|---|
| author | Li, Xiaohe Huang, Feilong Fan, Zide Mou, Fangli Hou, Yingyan Qian, Chen Wen, Lijie |
| author_facet | Li, Xiaohe Huang, Feilong Fan, Zide Mou, Fangli Hou, Yingyan Qian, Chen Wen, Lijie |
| contents | Trajectory prediction has garnered widespread attention in different fields, such as autonomous driving and robotic navigation. However, due to the significant variations in trajectory patterns across different scenarios, models trained in known environments often falter in unseen ones. To learn a generalized model that can directly handle unseen domains without requiring any model updating, we propose a novel meta-learning-based trajectory prediction method called MetaTra. This approach incorporates a Dual Trajectory Transformer (Dual-TT), which enables a thorough exploration of the individual intention and the interactions within group motion patterns in diverse scenarios. Building on this, we propose a meta-learning framework to simulate the generalization process between source and target domains. Furthermore, to enhance the stability of our prediction outcomes, we propose a Serial and Parallel Training (SPT) strategy along with a feature augmentation method named MetaMix. Experimental results on several real-world datasets confirm that MetaTra not only surpasses other state-of-the-art methods but also exhibits plug-and-play capabilities, particularly in the realm of domain generalization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_08221 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | MetaTra: Meta-Learning for Generalized Trajectory Prediction in Unseen Domain Li, Xiaohe Huang, Feilong Fan, Zide Mou, Fangli Hou, Yingyan Qian, Chen Wen, Lijie Robotics Computer Vision and Pattern Recognition Trajectory prediction has garnered widespread attention in different fields, such as autonomous driving and robotic navigation. However, due to the significant variations in trajectory patterns across different scenarios, models trained in known environments often falter in unseen ones. To learn a generalized model that can directly handle unseen domains without requiring any model updating, we propose a novel meta-learning-based trajectory prediction method called MetaTra. This approach incorporates a Dual Trajectory Transformer (Dual-TT), which enables a thorough exploration of the individual intention and the interactions within group motion patterns in diverse scenarios. Building on this, we propose a meta-learning framework to simulate the generalization process between source and target domains. Furthermore, to enhance the stability of our prediction outcomes, we propose a Serial and Parallel Training (SPT) strategy along with a feature augmentation method named MetaMix. Experimental results on several real-world datasets confirm that MetaTra not only surpasses other state-of-the-art methods but also exhibits plug-and-play capabilities, particularly in the realm of domain generalization. |
| title | MetaTra: Meta-Learning for Generalized Trajectory Prediction in Unseen Domain |
| topic | Robotics Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.08221 |