Towards Deep Active Learning in Avian Bioacoustics
This addresses the challenge of scarce annotations in passive acoustic monitoring for avian bioacoustics, but it appears incremental as it outlines an approach and pilot study without broad results.
The paper tackles the problem of adapting deep learning models to diverse environments in avian bioacoustics by proposing a deep active learning approach to reduce annotation costs, with a small-scale pilot study conducted.
Passive acoustic monitoring (PAM) in avian bioacoustics enables cost-effective and extensive data collection with minimal disruption to natural habitats. Despite advancements in computational avian bioacoustics, deep learning models continue to encounter challenges in adapting to diverse environments in practical PAM scenarios. This is primarily due to the scarcity of annotations, which requires labor-intensive efforts from human experts. Active learning (AL) reduces annotation cost and speed ups adaption to diverse scenarios by querying the most informative instances for labeling. This paper outlines a deep AL approach, introduces key challenges, and conducts a small-scale pilot study.