使用 BEVRestore 解决高分辨率鸟瞰图地图构建中的训练成本发散问题
Addressing Diverging Training Costs using BEVRestore for High-resolution Bird's Eye View Map Construction
AI总结：
针对高分辨率BEV地图构建中的训练成本发散问题，提出BEVRestore机制，通过将特征编码至低分辨率空间再恢复至高分辨率空间，实现内存高效且即插即用的地图构建。
摘要：
近年来，用于地图构建的 Bird's Eye View (BEV) 融合技术在城市环境建图方面展现出卓越效果。然而，其深层且庞大的架构会导致大量的反向传播内存和计算延迟。因此，该问题在构建高分辨率（HR）BEV 地图时构成了不可避免的瓶颈，因为其大尺寸特征会导致包括 GPU 内存消耗和计算延迟在内的成本显著增加，这被称为训练成本发散问题。受此问题影响，大多数现有方法采用低分辨率（LR）BEV，难以准确估计城市场景组件（如车道和人行道）的精确位置。由于这种不精确会导致危险的运动规划（如碰撞规避），因此必须解决训练成本发散问题。在本文中，我们利用新颖的 BEVRestore 机制来解决该问题。具体而言，我们提出的模型将每个传感器的特征编码到 LR BEV 空间，并将其恢复到 HR 空间，以建立一个内存高效的地图构建器。为此，我们引入了 BEV 恢复策略，该策略恢复了上采样 BEV 特征的混叠和块状伪影，并缩小了标签的宽度。我们的大量实验表明，所提出的机制提供了一个即插即用、内存高效的流程，能够实现具有广泛 BEV 范围的 HR 地图构建。
英文摘要：
Recent advancements in Bird's Eye View (BEV) fusion for map construction have demonstrated remarkable mapping of urban environments. However, their deep and bulky architecture incurs substantial amounts of backpropagation memory and computing latency. Consequently, the problem poses an unavoidable bottleneck in constructing high-resolution (HR) BEV maps, as their large-sized features cause significant increases in costs including GPU memory consumption and computing latency, named diverging training costs issue. Affected by the problem, most existing methods adopt low-resolution (LR) BEV and struggle to estimate the precise locations of urban scene components like road lanes, and sidewalks. As the imprecision leads to risky motion planning like collision avoidance, the diverging training costs issue has to be resolved. In this paper, we address the issue with our novel BEVRestore mechanism. Specifically, our proposed model encodes the features of each sensor to LR BEV space and restores them to HR space to establish a memory-efficient map constructor. To this end, we introduce the BEV restoration strategy, which restores aliasing, and blocky artifacts of the up-scaled BEV features, and narrows down the width of the labels. Our extensive experiments show that the proposed mechanism provides a plug-and-play, memory-efficient pipeline, enabling an HR map construction with a broad BEV scope.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
