外观
论文arXiv 2610.2290012 天前
面向越野动力学学习的结构化感知表示
原文链接：Structured Perception Representations for Off-road Dynamics Learning（arXiv 2610.22900 · 发布日期：2026-10-16）
Structured Perception Representations for Off-road Dynamics Learning
提出SpatialBEV利用预测状态周围BEV特征学习越野动力学，并在跨季节数据集和高速真机路线验证。
核心要点
- 高速越野自主导航需要准确控制，地形交互是车辆动力学的重要组成。
- SpatialBEV在每步查询预测状态周围的BEV特征，结合几何与语义线索预测下一状态。
- 作者还提出基准评测方法，用于挖掘数据集中最困难的场景。
- 在跨季节数据集上预测误差降低17.1%，真机最大横向误差降低25.3%。
验证范围与局限：验证包括跨季节数据集，以及速度≥10 m/s且几何上具有挑战的真机路线。
英文摘要 (Abstract)
Autonomously navigating off-road terrain at high speed demands accurate control, which often relies on dynamics models that capture complex interactions between the robot and terrain. In off-road regimes, terrain interaction is a major component of the vehicle's dynamics. Prior work attempts to capture this effect by training learned models of the vehicle's dynamics, conditioned on its perception of the local environment. However, said conditioning is often overly simplistic, ignoring the spatial structure of the dynamics problem. Our approach, SpatialBEV, exploits this spatial structure. At each step, we query the bird's-eye-view (BEV) features around the predicted state, conditioning the next prediction on current geometric and semantic cues. We additionally present a benchmarking approach that mines a dataset's most difficult scenarios. On datasets collected across seasons, SpatialBEV reduces prediction error by 17.1%, compared to physics-based and learned baselines. We further evaluate on hardware on high-speed ($\geq$10 m/s) and geometrically challenging courses with a 25.3% reduction in maximum cross-track error. More details can be found at https://theairlab.org/spatial-bev-dynamics .
作者：Micah Nye, Samuel Triest, Matthew Sivaprakasam, Isaiah Adu, Sebastian Scherer, Wenshan Wang
来源
本页事实对应的公开页面，以原文为准。