Williams Ben, van Merriënboer Bart, Dumoulin Vincent, Hamer Jenny, Fleishman Abram B, McKown Matthew, Munger Jill, Rice Aaron N, Lillis Ashlee, White Clemency, Hobbs Catherine, Razak Tries, Curnick David, Jones Kate E, Denton Tom
University College London, London, UK.
Institute of Zoology, Zoological Society of London, London, UK.
Philos Trans R Soc Lond B Biol Sci. 2025 Jun 12;380(1928):20240280. doi: 10.1098/rstb.2024.0280.
Machine learning has the potential to revolutionize passive acoustic monitoring (PAM) for ecological assessments. However, high annotation and computing costs limit the field's adoption. Generalizable pretrained networks can overcome these costs, but high-quality pretraining requires vast annotated libraries, limiting their current development to data-rich bird taxa. Here, we identify the optimum pretraining strategy for data-deficient domains, using tropical reefs as a representative case study. We assembled ReefSet, an annotated library of 57 000 reef sounds taken across 16 datasets, though still modest in scale compared to annotated bird libraries. We performed multiple pretraining experiments and found that pretraining on a library of bird audio 50 times the size of ReefSet provides notably superior generalizability on held-out reef datasets, with a mean area under the receiver operating characteristic curve (AUC-ROC) of 0.881 (±0.11), compared to pretraining on ReefSet itself or unrelated audio, with a mean AUC-ROC of 0.724 (±0.05) and 0.834 (±0.05), respectively. However, our key findings show that cross-domain mixing, where bird, reef and unrelated audio are combined during pretraining, provides superior transfer learning performance, with an AUC-ROC of 0.933 (±0.02). SurfPerch, our optimum pretrained network, provides a strong foundation for automated analysis of tropical reef and related PAM data with minimal annotation and computing costs.This article is part of the theme issue 'Acoustic monitoring for tropical ecology and conservation'.
机器学习有潜力彻底改变用于生态评估的被动声学监测（PAM）。然而，高注释和计算成本限制了该领域的应用。可通用的预训练网络可以克服这些成本，但高质量的预训练需要大量带注释的库，这限制了它们目前在数据丰富的鸟类分类群中的发展。在这里，我们以热带珊瑚礁为代表性案例研究，确定数据匮乏领域的最佳预训练策略。我们组装了ReefSet，这是一个包含从16个数据集中获取的57000个珊瑚礁声音注释库，不过与带注释的鸟类库相比，其规模仍然较小。我们进行了多次预训练实验，发现对一个比ReefSet大50倍的鸟类音频库进行预训练，在留出的珊瑚礁数据集上具有显著更高的通用性，接收器操作特征曲线（AUC-ROC）的平均面积为0.881（±0.11），而在ReefSet本身或无关音频上进行预训练时，AUC-ROC的平均面积分别为0.724（±0.05）和0.834（±0.05）。然而，我们的主要发现表明，跨域混合，即在预训练期间将鸟类、珊瑚礁和无关音频组合起来，提供了更好的迁移学习性能，AUC-ROC为0.933（±0.02）。我们的最佳预训练网络SurfPerch为热带珊瑚礁及相关PAM数据的自动分析提供了坚实基础，所需注释和计算成本最低。本文是主题为“热带生态与保护的声学监测”的一部分。
