The development of a convolutional neural network for the automatic detection of Northern Bobwhite Colinus virginianus covey calls

Abstract

Passive acoustic monitoring using Autonomous Recording Units (ARUs) is becoming a significant research tool for collecting large amounts of ecological data. Northern bobwhite Colinus virginianus is an economically important game bird whose declining populations are of conservation concern, so efforts to monitor bobwhite abundance using ARUs are being intensified. Yet, manual processing of ARU data is time consuming and often expensive, so developing automatic call detection methods is a key step in acoustic monitoring. We present here the first single species convolutional neural network (CNN) developed purely for automatic bobwhite covey call identification and classification. We demonstrate the value of meaningful data augmentation by including nontarget calls and background noise into our training dataset, as well as evaluating alternative CNN score thresholds and model extrapolation performance. We trained our CNN on 6,682 manually labeled covey calls across three groups of sites within the southeastern USA. Precision and AUC from both CNN classification and individual call detection was high (0.80-0.99), and our model showed strong extrapolation ability across site groups. However, extrapolation performance significantly decreased for sites that were more dissimilar to the training data set if our meaningful data augmentation process was omitted. Our CNN detected significantly more covey calls than manual labeling using Raven Pro software, and processing time was greatly reduced: a single one hour wav file can be now analyzed by the CNN in roughly eight seconds. We also demonstrate using a simple case study that extremely high variability in estimates of bobwhite site occupancy and detection are obtained depending on the method of acoustic data processing (manual versus CNN). Our results suggest that our CNN provides robust and time-saving analysis of bobwhite covey call acoustic data and can be applied to future research and monitoring projects with high confidence in the performance of the model.

An efficient acoustic classifier for high‐priority avian species in the southern Great Plains using convolutional neural networks

“…Paumen et al 2022). When compared to other avian studies (Ruff et al 2020, Gillings and Scott 2021, Nolan et al 2022, our trained multilabel CNN achieved similar or even higher levels of classification accuracy ( ≥ 99%) and proved to be highly capable of distinguishing between the 5 different avian species calls herein. The CNN also successfully identified portions of recordings that were devoid of bird calls (i.e., noise clips), resulting in a false positive error rate of zero.…”

“…We chose to use a pretrained multilabel classifier, the multilabel vgg16 architecture, to Passive acoustic monitoring data collected using ARUs are becoming an increasingly common input to population and species distribution models (Hagens et al 2018, Sebastián-González et al 2018, Stevenson et al 2021. Models estimating distribution, occurrence, abundance, or density have been fitted using ARU data from a wide variety of vocalizing taxa (Marques et al 2013, Sebastián-González et al 2018, Nolan et al 2022 and have been shown to outperform traditional survey methods such as human observer point counts (Alquezar and Machado 2015). Convolutional neural networks and other automated detection algorithms therefore have great potential to facilitate the expansion of ARU deployment and recording across geographic space and time.…”

OpenSoundscape: An open‐source bioacoustics analysis package for Python

“…When research groups have developed domain‐specific deep learning classifiers, automated classifier performance can be excellent and can lead to novel biological insights (e.g. Bermant et al, 2019; Nolan et al, 2023; Wightman et al, 2022; Zhong et al, 2021). Making such analysis methods more accessible to researchers could broadly improve the quality of bioacoustic data analyses and enhance insights into ecological processes (Stowell, 2022).…”

Optimizing detection of the bobwhite reproduction call using passive acoustic monitoring

“…The decline of northern bobwhite populations across much of its range has resulted in increased monitoring efforts and interest in capitalizing on new technologies such as ARUs (Brennan 1991, Hernandez et al 2013, Wilhite et al 2020, Nolan et al 2023). One subspecies of northern bobwhite, the endangered masked bobwhite ( C. v. ridgwayi ), suffered declines shortly after its discovery in 1884 and was extirpated on the northern end of its distribution in south‐central Arizona, USA, by 1900 (Breninger 1904, Brown 1904).…”

“…attenuation, Colinus virginianus, Kaleidoscope, masked bobwhite, sound, whistle count Surveys based on detections of vocalizations are commonly used to monitor avian populations. For example, surveys based on detection of the northern bobwhite (Colinus virginianus) reproduction call (i.e., spring whistle count) are used to index breeding populations and draw inferences about relative abundance, occupancy, and trends, and predict autumn abundance, measure response to management actions, and track nesting chronology among others (Terhune et al 2009, Sisson and Terhune 2017, Nolan et al 2023. For some applications, the timing of spring whistle counts can provide advantages over autumn covey call counts and have been used to index breeding populations of bobwhite for >90 years (Stoddard 1931).…”