提出线性内存的SE(2)不变注意力，提升自动驾驶多智能体建模效率。
Linear Memory SE(2) Invariant Attention
- 基于相对位姿设计线性内存的SE(2)不变注意力机制
- 在多个数据集上性能优于非不变架构，且内存开销仅线性增长
- 适合需要高效处理多智能体场景的自动驾驶系统研发
自动驾驶中的运动预测、多智能体模拟和规划等任务依赖于空间数据处理。已有方法采用考虑物体间相对位姿的SE(2)不变网络结构，但需显式计算所有物体对的相对位姿，导致内存开销为二次方。本文提出一种SE(2)不变的缩放点积注意力机制，内存复杂度随物体数量线性增长。所提出的SE(2)不变Transformer架构具备类似大语言模型的可扩展性。实验表明，该方法实现可行且相比同类非不变架构性能更优。
原文摘要 · Abstract (English)
Processing spatial data is a key component in many learning tasks for autonomous driving such as motion forecasting, multi-agent simulation, and planning. Prior works have demonstrated the value in using SE(2) invariant network architectures that consider only the relative poses between objects (e.g. other agents, scene features such as traffic lanes). However, these methods compute the relative poses for all pairs of objects explicitly, requiring quadratic memory. In this work, we propose a mechanism for SE(2) invariant scaled dot-product attention that requires linear memory relative to the number of objects in the scene. Our SE(2) invariant transformer architecture enjoys the same scaling properties that have benefited large language models in recent years. We demonstrate experimentally that our approach is practical to implement and improves performance compared to comparable non-invariant architectures.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。