发表机构
Wayve Technologies（Wayve Technologies）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
BLInD利用车辆状态历史学习未来自我轨迹的top-k分布，无需传感器输入，实现低延迟跨域泛化，并显著降低AEB误报率。
AI 中文摘要
我们提出了BLInD（盲学习意图分布），一个紧凑的网络，将最近的车辆状态历史（例如速度、曲率、转向灯和车辆类型）映射到未来自我轨迹的top-k分布，无需摄像头、激光雷达、地图或物体跟踪输入。我们发现仅车辆状态历史就足以学习近端自我轨迹的有用多模态分布，其低延迟特性使其非常适合安全关键的部署。我们研究了两种分布架构，自回归（AR）和流匹配，并在混合平台开源数据集和Wayve数据集上训练。两者均无需数据集特定适配即可泛化；流匹配模型在Wayve上达到最佳top-k ADE/FDE为0.15/0.37米，在Waymo上为0.15/0.36米，在nuScenes上为0.28/0.59米，AR模型达到相当的覆盖率。将分布集成到AEB触发任务中，与100%真正例得分的1-CTRV策略相比，严格的全候选策略将误报率从1.51%降至0.11%（AR，13.7倍减少，94.9%真正例率）和0.06%（流匹配，25.1倍减少，98.7%真正例率）。BLInD在NVIDIA DRIVE Orin ECU上以AR头0.87毫秒和流匹配头2.9毫秒运行，使其与汽车ECU上的实时部署兼容。虽然现有的学习分布模型依赖场景上下文，且盲车辆状态模型通常坍缩为单一路径，BLInD同时具备学习、盲和跨域特性，这是先前工作未展示的组合。这些结果表明，这种分布为下游系统提供了可控且合理的意图采样接口，其中AEB是一个实例。
英文摘要
We present BLInD (Blind Learned Intent Distribution), a compact network that maps recent vehicle-state history (e.g. speed, curvature, indicator, and vehicle type) to a top-k distribution of future ego trajectories, with no camera, LiDAR, map, or object-track inputs. We find that vehiclestate history alone is sufficient to learn a useful multimodal distribution over near-term ego trajectories, and its low-latency nature makes it well-suited for safety-critical deployment. We investigate two distribution architectures, autoregressive (AR) and flow-matching, and train on both mixed-platform opensource and Wayve datasets. Both generalize without datasetspecific adaptation; the flow-matching model achieves best topk ADE/FDE of 0.15/0.37 m on Wayve, 0.15/0.36 m on Waymo, and 0.28/0.59 m on nuScenes, with the AR model reaching comparable coverage. Integrating the distributions into an AEB trigger task, a strict all-candidates policy reduces false positives from 1.51% to 0.11% with AR (13.7x reduction, 94.9% TP) and to 0.06% with flow-matching (25.1x reduction, 98.7% TP) compared to a 1-CTRV policy with 100% true positive score. BLInD runs in 0.87 ms with the AR head and 2.9 ms with the flow-matching head on an NVIDIA DRIVE Orin ECU making it compatible with real-time deployment on automotive ECUs. While existing learned distribution models rely on scene context and blind vehicle-state models typically collapse to a single path, BLInD is learned, blind, and cross-domain simultaneously, a combination not demonstrated by prior work. These results show that such a distribution provides a controllable and plausible intent sampling interface for downstream systems, with AEB as one instantiation.
CommentsAccepted at the 2026 IEEE International Conference on Intelligent Transportation Systems (ITSC)
