发表机构
Fondazione Bruno Kessler; University of Trento; University College London（布鲁诺·凯斯勒基金会; 特伦托大学; 伦敦大学学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对蝙蝠被动声学监测的数据洪流问题，提出硬件感知的BAD检测器，在微控制器上实现100%硬件卸载，以高精度区分蝙蝠叫声与干扰，显著提升精度。
AI 中文摘要
对蝙蝠进行被动声学监测会产生海量超声波数据集（每个节点每晚超过27 GB），给边缘存储和电池寿命带来压力。传统触发器无法应对声学干扰源，而深度模型又超出微控制器的限制。我们提出了一种硬件感知的蝙蝠活动检测器（BAD），专门用于在可变采样率（192-384 kHz）下区分蝙蝠叫声与困难的生物和环境干扰源。针对Silicon Labs EFM32PG26（MVP）以8位整数精度定制，我们的模型在全部14层（17.2 KB闪存，73.1 KB RAM）实现了100%的硬件卸载。端到端预处理（76帧耗时74.00 ms）和推理（30.00 ms）在192 kHz下处理100 ms片段需104.00 ms。在现实低流行率（r_pos = 0.05）条件下的空间域外录音上，BAD实现了0.9748的AUC-ROC，抑制了99.4%的非目标噪声帧，同时保留了65.3%的蝙蝠叫声——相比经典Goertzel基线，精度提升了超过33倍。
英文摘要
Passive Acoustic Monitoring of bats generates massive ultrasonic datasets (>27 GB/night per node), straining edge storage and battery life. Legacy triggers fail against acoustic confusers, while deep models exceed microcontroller limits. We present a hardware-aware Bat Activity Detector (BAD) specifically designed to discriminate bat calls from hard biological and environmental confusers across variable sampling rates (192-384 kHz). Tailored for the Silicon Labs EFM32PG26 (MVP) in 8-bit integer precision, our model achieves 100 percent hardware offload across all 14 layers (17.2 KB Flash, 73.1 KB RAM). End-to-end preprocessing (74.00 ms for 76 frames) and inference (30.00 ms) of 100 ms clips at 192 kHz require 104.00 ms per clip. On spatially out-of-domain recordings under a realistic low-prevalence regime (r_pos = 0.05), BAD achieves an AUC-ROC of 0.9748 and suppresses 99.4% of non-target noise frames while retaining 65.3% of bat calls - delivering a >33x precision gain over classical Goertzel baselines.
