密集覆盖，稀疏细化：字节受限的协同感知
Dense Coverage, Sparse Refinement: Byte-Constrained Cooperative Perception
- FZI Research Center for Information Technology（FZI信息技术研究中心）
- Karlsruhe Institute of Technology（卡尔斯鲁厄理工学院）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
针对V2X带宽限制下的协同感知，提出覆盖-细化方法：全图粗层加高价值补丁细化，任务感知选择器分配预算，在DAIR-V2X和OPV2V上以千字节级负载实现高精度。
AI中文摘要：
协同感知通过跨连接智能体共享中间鸟瞰图（BEV）特征来提升自动驾驶感知性能，但在严格的车联万物（V2X）带宽限制下，密集特征交换难以部署。现有高效方法通常要么对完整特征图进行均匀压缩，将比特花费在低价值背景上，要么稀疏化通信，冒着丢失有用上下文的风险。我们提出了一种用于字节受限协同感知的覆盖-细化设计：每个智能体在全BEV图上传输高度压缩的粗层，并将剩余预算分配给选定的高分辨率补丁。一个任务感知效益选择器根据估计的下游效用对单元进行排序，实现确定性的预算细化以及对带宽变化的零重训练适应。接收端重建与标准融合模块兼容的密集BEV张量。在DAIR-V2X和OPV2V上的实验显示了在千字节级预算下的强精度-负载权衡。在DAIR-V2X上，我们的方法在每非本车智能体仅1.87 KB时达到0.60 AP@0.7，而均匀SimVQ压缩在4.61 KB时为0.52。受控诊断进一步表明，增益来自覆盖-细化分配而非仅量化。代码将发布。
英文摘要：
Collaborative perception improves autonomous perception by sharing intermediate Bird's-Eye-View (BEV) features across connected agents, but dense feature exchange is difficult to deploy under strict Vehicle-to-Everything (V2X) bandwidth limits. Existing efficient methods typically either compress the full feature map uniformly, spending bits on low-value background, or sparsify communication, risking the loss of useful context. We propose a coverage-refinement design for byte-constrained cooperative perception: each agent transmits a highly compressed coarse layer over the full BEV map and allocates the remaining budget to selected high-resolution patches. A Task-Aware Benefit Selector ranks cells by estimated downstream utility, enabling deterministic budgeted refinement and zero-retraining adaptation to changing bandwidth. The receiver reconstructs a dense BEV tensor compatible with standard fusion modules. Experiments on DAIR-V2X and OPV2V show strong accuracy-payload trade-offs at kilobyte-scale budgets. On DAIR-V2X, our method reaches 0.60 AP@0.7 at only 1.87 KB per non-ego agent, compared with 0.52 at 4.61 KB for uniform SimVQ compression. Controlled diagnostics further show that the gain arises from coverage-refinement allocation rather than quantization alone. Code will be published.
