Automated Bioacoustic Monitoring for South African Bird Species on Unlabeled Data

TL;DR

This paper presents an automated framework that extracts labeled bird sound data from unlabeled recordings, trains CRNN models for bird detection, and demonstrates effective performance in noisy urban environments, aiding biodiversity monitoring.

Contribution

The study introduces a novel method for automatically generating labeled training data from unlabeled recordings, enabling adaptable bird sound detection models without extensive manual labeling.

Findings

CRNN model achieved F1 score of 0.73 in real-world noisy data

Automated data extraction simplifies adaptation to new species and habitats

Framework enhances passive acoustic monitoring for conservation efforts

Abstract

Analyses for biodiversity monitoring based on passive acoustic monitoring (PAM) recordings is time-consuming and challenged by the presence of background noise in recordings. Existing models for sound event detection (SED) worked only on certain avian species and the development of further models required labeled data. The developed framework automatically extracted labeled data from available platforms for selected avian species. The labeled data were embedded into recordings, including environmental sounds and noise, and were used to train convolutional recurrent neural network (CRNN) models. The models were evaluated on unprocessed real world data recorded in urban KwaZulu-Natal habitats. The Adapted SED-CRNN model reached a F1 score of 0.73, demonstrating its efficiency under noisy, real-world conditions. The proposed approach to automatically extract labeled data for chosen avian species enables an easy adaption of PAM to other species and habitats for future conservation projects.