Real-Time Motion Prediction via Heterogeneous Polyline Transformer with Relative Pose Encoding
![Thumbnail](/bitstream/handle/20.500.11850/643424/8148_real_time_motion_prediction_vi.pdf.jpg?sequence=5&isAllowed=y)
Open access
Autor(in)
Alle anzeigen
Datum
2024-07Typ
- Conference Paper
ETH Bibliographie
yes
Altmetrics
Abstract
The real-world deployment of an autonomous driving system requires its components to run on-board and in real-time, including the motion prediction module that predicts the future trajectories of surrounding traffic participants. Existing agent-centric methods have demonstrated outstanding performance on public benchmarks. However, they suffer from high computational overhead and poor scalability as the number of agents to be predicted increases. To address this problem, we introduce the K-nearest neighbor attention with relative pose encoding (KNARPE), a novel attention mechanism allowing the pairwise-relative representation to be used by Transformers. Then, based on KNARPE we present the heterogeneous polyline Transformer with relative pose encoding (HPTR), a hierarchical framework enabling asynchronous token update during the online inference. By sharing contexts among agents and reusing the unchanged contexts, our approach is as efficient as scene-centric methods, while performing on par with state-of-the-art agent-centric methods. Experiments on Waymo and Argoverse-2 datasets show that HPTR achieves superior performance among end-to-end methods that do not apply expensive post-processing or ensembling. The code is available at https://github.com/zhejz/HPTR. Mehr anzeigen
Persistenter Link
https://doi.org/10.3929/ethz-b-000643424Publikationsstatus
publishedExterne Links
Herausgeber(in)
Buchtitel
Advances in Neural Information Processing Systems 36Seiten / Artikelnummer
Verlag
CurranKonferenz
Organisationseinheit
03514 - Van Gool, Luc / Van Gool, Luc
09688 - Yu, Fisher / Yu, Fisher
Zugehörige Publikationen und Daten
Is supplemented by: https://github.com/zhejz/HPTR
Is new version of: https://doi.org/10.48550/ARXIV.2310.12970
Anmerkungen
Conference lecture on December 13, 2023.ETH Bibliographie
yes
Altmetrics