研究人员开发了 DenseScout，一种用于边缘设备高效小目标选择的新型算法-系统协同设计方法。这种轻量级选择器仅包含 101 万个参数，在严格的计算和延迟限制下，比传统的基于检测器的方法更能有效地优先选择候选图像块。该系统还采用了感知传输的运行时实现和 QoS 约束召回率指标，以确保在异构硬件上满足截止时间内的性能。 AI
影响 通过整合算法和系统设计，优化了资源受限边缘设备上的小目标感知。
排序理由 这是一篇研究论文，详细介绍了用于边缘平台目标选择的新算法和系统协同设计。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
Deploying tiny object perception on edge platforms is challenging because practical systems must satisfy both strict compute budgets and end-to-end latency constraints. A common strategy is to first select a small number of candidate patches from a high-resolution image and then …
arXiv:2604.25300v1 Announce Type: new Abstract: Deploying tiny object perception on edge platforms is challenging because practical systems must satisfy both strict compute budgets and end-to-end latency constraints. A common strategy is to first select a small number of candidat…
Deploying tiny object perception on edge platforms is challenging because practical systems must satisfy both strict compute budgets and end-to-end latency constraints. A common strategy is to first select a small number of candidate patches from a high-resolution image and then …