SparseAD: Sparse Query-Centric Paradigm for Efficient End-to-End . . . In this paper, we propose a Sparse query-centric paradigm for end-to-end Autonomous Driving (SparseAD), where the sparse queries completely represent the whole driving scenario across space, time and tasks without any dense BEV representation
GitHub - swc-17 SparseDrive: SparseDrive: End-to-End Autonomous Driving . . . SparseDrive is a Sparse-Centric paradigm for end-to-end autonomous driving We explore the sparse scene representation for end-to-end autonomous driving and propose a Sparse-Centric paradigm named SparseDrive, which unifies multiple tasks with sparse instance representation
[2404. 06892] SparseAD: Sparse Query-Centric Paradigm for . . . - ar5iv In this paper, we propose a novel Sparse Query-Centric paradigm for end-to-end Autonomous Driving (SparseAD), in which all elements across space and time in the whole driving scenario are represented by sparse queries without any dense BEV features, as shown in Fig 2 (c)
SparseAD: Sparse Query-Centric Paradigm for Efficient End . . . 本论文旨在提出一种Sparse query-centric paradigm用于end-to-end自动驾驶，以解决现有end-to-end方法在子任务上表现不佳的问题。 同时，该方法不需要使用昂贵的dense BEV features，可以更容易地扩展到更多模态或任务。 SparseAD采用了稀疏查询的方式，完全代表了整个驾驶场景，包括空间、时间和任务，而不需要任何密集的BEV表示。 同时，设计了一个统一的稀疏架构，用于感知任务，包括检测、跟踪和在线地图制作。 此外，重新审视了运动预测和规划，并设计了一个更合理的运动规划框架。