Rules of the Road: Predicting Driving Behavior with a Convolutional Model of Semantic Interactions
Joey Hong Caltech Work done while at Zoox. The authors would also like to thank and acknowledge Kai Wang for his work on this project. Kai has been instrumental in coordinating the final version of the paper and preparing the dataset for release. Benjamin Sapp James Philbin Zoox
Abstract
We focus on the problem of predicting future states of entities in complex, real-world driving scenarios. Previous research has used low-level signals to predict short time horizons, and has not addressed how to leverage key assets relied upon heavily by industry self-driving systems: (1) large 3D perception efforts which provide highly accurate 3D states of agents with rich attributes, and (2) detailed and accurate semantic maps of the environment (lanes, traffic lights, crosswalks, etc). We present a unified representation which encodes such high-level semantic information in a spatial grid, allowing the use of deep convolutional models to fuse complex scene context. This enables learning entity-entity and entity-environment interactions with simple, feed-forward computations in each timestep within an overall temporal model of an agent’s behavior. We propose different ways of modelling the future as a distribution over future states using standard supervised learning. We introduce a novel dataset providing industry-grade rich perception and semantic inputs, and empirically show we can effectively learn fundamentals of driving behavior.
中文速览
自动驾驶系统需要准确预测周围车辆未来几秒的行驶轨迹，但现有方法要么只用原始传感器信号、预测时间太短，要么忽略了行业实际依赖的两大核心资产：高精度3D感知结果和详细语义地图。本文提出将这些高层语义信息（车道、交通灯、人行横道等）与动态实体状态一起渲染成统一的俯视空间网格，再用深度卷积网络一次性前向推理，同时建模车辆自身动态、车辆间交互和环境上下文。在预测未来状态时，模型不输出单一轨迹，而是输出多模态概率分布（高斯混合或离散占用网格），以覆盖车辆可能采取的多种行为意图。为此，作者还发布了一个规模远超现有公开数据集的新数据集（88个地点、近17万小时等效场景、含语义地图），实验证明模型能有效学习驾驶行为规律，预测精度在5秒时间跨度内显著优于基线。这项工作填补了工业级感知与预测研究之间的鸿沟，为自动驾驶规划系统提供了更实用、更安全的多模态轨迹预测方案。
原文 arXiv:1906.08945;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1906.08945v1