用神经形态芯片实现低功耗持续声学异常检测
Low-Power, Neuromorphic, Acoustic Anomaly Detection for Persistent Machine Monitoring
- 用英特尔Loihi 2芯片运行自编码器，离线提取梅尔特征后本地推理
- 在无噪环境下达到0.9959的AUC，噪声环境下仍超基线表现
- 每样本仅需0.0426毫焦能耗，比CPU/GPU低两个数量级
持续声学监测可在不接触设备的情况下检测机械故障，但始终在线推理受限于功耗、延迟和部署复杂性。本文在英特尔Loihi 2神经形态处理器上实现了基于自编码器的声学异常检测，在干净与噪声条件下均有效。梅尔对数特征在芯片外计算；归一化、自编码器推理、L1重构评分与阈值判定均在芯片内完成。在无噪、麦克风位置无关的ToyADMOS ToyCar基准测试中，芯片内模型达到0.9959 AUC与0.9785标准化pAUC（最大假阳性率0.1）。在DCASE 2026 Task 2 ToyCar噪声基准测试中，源域AUC为0.7990，目标域AUC为0.6466，pAUC为0.6426，超过报告基线。在16芯片Loihi 2 VPX系统上的功耗分析显示，实时吞吐下每样本动态能量为0.0406–0.0426毫焦，较CPU与GPU低两个数量级。这些结果表明神经形态声学异常检测是低功耗持续机器监控的可行方案。
原文摘要 · Abstract (English)
Persistent acoustic monitoring can detect machine faults without physical contact, but always-on inference is constrained by power, latency, and deployment complexity. We demonstrate autoencoder-based acoustic anomaly detection on an Intel Loihi 2 neuromorphic processor under clean and noisy conditions. Log-mel features are computed off chip; normalization, autoencoder inference, L1 reconstruction scoring, and thresholding run on chip. In a clean, microphone-position-invariant ToyADMOS ToyCar benchmark, the on-chip model achieves 0.9959 AUC and 0.9785 standardized pAUC at maximum false-positive rate 0.1. In the DCASE 2026 Task 2 ToyCar noisy benchmark, the model achieves source AUC 0.7990, target AUC 0.6466, and pAUC 0.6426, exceeding reported baseline metrics. Power profiling on a 16-chip Loihi 2 VPX system shows real-time throughput with 0.0406$\unicode{x2013}$0.0426 mJ dynamic energy per sample, two orders of magnitude lower than both a CPU and GPU. These results support neuromorphic acoustic anomaly detection as a practical candidate for low-power, persistent machine monitoring.
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。