Bad: Taming the Bioacoustic Data Deluge with a Bat Acticity Detector

2026-09-29

AI Summary

This study addresses the challenges of limited storage and power on edge devices, weak interference robustness in conventional methods, and insufficient computational capacity for deep models in passive acoustic monitoring. We propose a hardware-aware bat call detector tailored for microcontrollers. The core innovation lies in the first end-to-end preprocessing and inference architecture supporting 8-bit integer precision, which achieves full-layer hardware offloading via a 14-layer neural network on the Silicon Labs EFM32PG26 while accommodating variable sampling rates from 192 to 384 kHz. Experimental results demonstrate that the proposed model attains an AUC of 0.9748, suppresses 99.4% of environmental noise while preserving 65.3% of target signals, and outperforms traditional baselines by over 33-fold in accuracy, effectively unifying high detection performance with ultra-low power consumption under extremely constrained resources.

Abstract

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.

Problem

• Passive Acoustic Monitoring
• Bioacoustics
• Edge Computing
• Bat Activity Detection
• TinyML

Innovation

• Hardware-aware Neural Network
• Edge Computing
• Bioacoustic Monitoring
• TinyML
• Bat Activity Detection