TinyML实时不确定性监控，无需标签和额外计算
TCUQ: Single-Pass Uncertainty Quantification from Temporal Consistency with Streaming Conformal Calibration for TinyML
- 利用短期时序一致性生成校准风险分数，仅需常数时间更新
- 在微控制器上内存减小50%-60%，速度提升30%-45%，准确率损失检测提升3-7个AUPRC点
- 适合资源受限设备的在线监控，特别适用于无标签数据流
我们提出TCUQ，一种针对流式TinyML的单遍、无标签不确定性监控方法。该方法通过轻量级后验与特征信号捕捉短时序一致性，结合O(W)环形缓冲区与O(1)每步更新，生成可校准的风险分数。一个流式共形层将该分数转化为有预算的接受/弃权规则，实现校准行为而无需在线标签或额外前向传播。在微控制器上，TCUQ可在千字节级设备上运行，相比早退和深度集成方案，内存减少约50%至60%，延迟降低约30%至45%；相似精度方法常因内存不足无法运行。在分布内数据流受扰场景下，TCUQ使准确率下降检测提升3至7 AUPRC点，高严重性下可达0.86 AUPRC；故障检测达到最高0.92 AUROC。结果表明，时序一致性结合流式共形校准，为TinyML设备端监控提供了高效实用的基础。
原文摘要 · Abstract (English)
We introduce TCUQ, a single pass, label free uncertainty monitor for streaming TinyML that converts short horizon temporal consistency captured via lightweight signals on posteriors and features into a calibrated risk score with an O(W ) ring buffer and O(1) per step updates. A streaming conformal layer turns this score into a budgeted accept/abstain rule, yielding calibrated behavior without online labels or extra forward passes. On microcontrollers, TCUQ fits comfortably on kilobyte scale devices and reduces footprint and latency versus early exit and deep ensembles (typically about 50 to 60% smaller and about 30 to 45% faster), while methods of similar accuracy often run out of memory. Under corrupted in distribution streams, TCUQ improves accuracy drop detection by 3 to 7 AUPRC points and reaches up to 0.86 AUPRC at high severities; for failure detection it attains up to 0.92 AUROC. These results show that temporal consistency, coupled with streaming conformal calibration, provides a practical and resource efficient foundation for on device monitoring in TinyML.
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。