# Bioacoustic Detection of Wolves Using AI (BirdNET, Cry-Wolf and BioLingual)

## TL;DR

This study shows that AI can detect wolf howls in audio recordings, offering a faster and non-invasive way to monitor wolf populations.

## Contribution

The study introduces a combined AI approach that significantly improves detection rates of wolf howls compared to individual AI methods.

## Key findings

- BirdNET detected 78.5% of wolf howls but had many false positives.
- Combining BirdNET, BioLingual, and Cry-Wolf achieved 96.2% detection of howls.
- AI methods reduced analysis time, making large-scale monitoring feasible.

## Abstract

Assessment of wolf (Canis lupus) populations today relies on multiple time-consuming and resource-intensive methods, including DNA testing of feces and wolf kills and wolf observations on wildlife cameras. This study aimed to explore wolf howls as an alternative monitoring tool for wolves and to compare several Artificial Intelligence (AI) methods against a baseline of manual registration for detecting and classifying wolf howls from audio recordings. The results show that AI-based methods like BirdNET, BioLingual, and Cry-Wolf achieved high detection rates (78.5%, 61.5%, and 59.6% recall—the proportion of actual howls successfully detected—respectively), though they also produced a substantial number of false positives. Crucially, combining these AI methods yielded an impressive 96.2% recall for actual howls. The use of automated AI methods significantly reduced the time spent on analysis of recordings, enabling the processing of larger datasets with fewer resources. This study demonstrates how the integration of AI-driven acoustic analysis can act as a non-invasive and efficient method, holding the possibility of becoming a standard for monitoring wolf populations and many other animal species.

Rising numbers of wolf (Canis lupus) populations make traditional, resource-intensive methods of wolf monitoring increasingly challenging and often insufficient. This study explores how wolf howls can be used as a new monitoring tool for wolves by applying Artificial Intelligence (AI) methods to detect and classify wolf howls from acoustic recordings, thereby improving the effectiveness of wolf population monitoring. Three AI approaches are evaluated: BirdNET, Yellowstone’s Cry-Wolf project system, and BioLingual. Data were collected using Song Meter SM4 (SM4) audio recorders in a known wolf territory in Klelund Dyrehave, Denmark, and manually validated to establish a ground truth of 260 wolf howls. Results demonstrate that while AI solutions currently do not achieve the complete precision or overall accuracy of expert manual analysis, they offer tremendous efficiency gains, significantly reducing processing time. BirdNET achieved the highest recall at 78.5% (204/260 howls detected), though with a low precision of 0.007 (resulting in 28,773 false positives). BioLingual detected 61.5% of howls (160/260) with 0.005 precision (30,163 false positives), and Cry-Wolf detected 59.6% of howls (155/260) with 0.005 precision (30,099 false positives). Crucially, a combined approach utilizing all three models achieved a 96.2% recall (250/260 howls detected). This suggests that while AI solutions primarily function as powerful human-aided data reduction tools rather than fully autonomous detectors, they represent a valuable, scalable, and non-invasive complement to traditional methods in wolf research and conservation, making large-scale monitoring more feasible.