TERRA：面向任务的边缘 AI 区域火灾检测与无身份报告
TERRA: Task-Oriented Edge AI for Regional Fire Detection with Identity-Free Reporting
Mohammad Kazemi · Julie A. McCann
中文摘要
分布式火灾监测系统可能包含大量传感设备，通过共享上行链路报告，但运行目标通常是识别有火灾证据的区域，而非恢复设备级测量。提出 TERRA（Task-Oriented Edge Representation and Regional Aggregation），一种边缘 AI 框架，结合面向任务的本地推理、无身份区域标记报告和置换不变的区域证据聚合。TERRA 支持两种报告模式：TERRA-Score 中每个设备发送量化后的本地火灾概率；TERRA-Feature 中设备发送紧凑学习表示，接收端用置换不变的 DeepSets 模型聚合。以分布式室内 EN54 火灾报警数据集的全部 15 种数值传感模态为概念验证测试平台，在重复不相交的三折交叉验证下评估面向事件级火灾场景目标的区域推理。研究表示大小、任务载荷量化、报告可用性以及对漏报和误恢复消息的鲁棒性。在 9 条报告且消息完美恢复时，TERRA-Feature 使用 24 位任务载荷达到平均 AUPRC 0.990，TERRA-Score 仅用 6 位达到 0.987；两者在 FAR ≤1% 时 MDR 均为 1.81%。可靠恢复下 TERRA-Feature 的 AUPRC 略高，而更紧凑的 TERRA-Score 对虚假消息污染增加更不敏感。
关键要点
- 01问题：分布式火灾监测需识别区域火灾证据，而非恢复设备级测量，共享上行链路资源受限。
- 02方法：TERRA 结合本地任务推理、无身份区域标记报告与置换不变 DeepSets 聚合。
- 03结果：9 条报告完美恢复时，TERRA-Feature 24 位载荷 AUPRC 0.990，TERRA-Score 6 位达 0.987。
- 04结果：两者在 FAR ≤1% 时 MDR 均为 1.81%，TERRA-Score 对虚假消息污染更鲁棒。
解读
尚无解读。
原始英文摘要
arXiv:2610.37909v1 Announce Type: cross Abstract: Distributed fire-monitoring systems may involve many sensing devices reporting over a shared uplink, even though the operational objective is often to identify regions exhibiting evidence of fire rather than to recover device-specific measurements. We introduce TERRA (Task-Oriented Edge Representation and Regional Aggregation), an edge-AI framework that combines task-oriented local inference, identity-free region-tagged reporting, and permutation-invariant regional evidence aggregation. TERRA supports two reporting modes: TERRA-Score, in which each device communicates a quantized local fire probability, and TERRA-Feature, in which it communicates a compact learned representation for receiver-side aggregation using a permutation-invariant DeepSets model. Using all 15 numerical sensing modalities of the distributed indoor EN54 fire-alarm dataset as a proof-of-concept testbed, we evaluate regional inference for an episode-level fire-scenario target under repetition-disjoint three-fold cross-validation. We study representation size, task-payload quantization, report availability, and robustness to missed and false recovered messages. With nine reports and perfect message recovery, TERRA-Feature achieves mean AUPRC 0.990 using a 24-bit task payload, while TERRA-Score achieves 0.987 using only 6 bits; both attain an MDR of 1.81% at FAR $\leq$1%. TERRA-Feature achieves slightly higher AUPRC under reliable recovery, whereas the more compact TERRA-Score is less sensitive to increasing false-message contamination.