论文标题
感知，参加和开车：学习空间关注以进行安全自动驾驶
Perceive, Attend, and Drive: Learning Spatial Attention for Safe Self-Driving
论文作者
论文摘要
在本文中，我们提出了一个端到端的自动驾驶网络，该网络具有一个稀疏的注意模块，该模块学会自动参加输入的重要区域。注意模块专门针对运动计划，而先前的文献仅在感知任务中应用了注意力。学习直接针对运动计划的注意力面罩可以通过执行更集中的计算来显着提高计划者的安全性。此外，可视化注意力可以提高端到端自动驾驶的解释性。
In this paper, we propose an end-to-end self-driving network featuring a sparse attention module that learns to automatically attend to important regions of the input. The attention module specifically targets motion planning, whereas prior literature only applied attention in perception tasks. Learning an attention mask directly targeted for motion planning significantly improves the planner safety by performing more focused computation. Furthermore, visualizing the attention improves interpretability of end-to-end self-driving.
