Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study introduces MetaPerch, a bioacoustic foundation model that leverages auxiliary metadata such as location and time to enhance species detection performance. By integrating metadata as additional supervision signals, the model captures species-metadata correlations, resulting in a more robust representation that generalizes effectively across various ecological contexts. The empirical analysis demonstrates that utilizing nine different metadata sources significantly improves identification accuracy across 17 diverse bioacoustic datasets, addressing critical challenges in real-world passive acoustic monitoring.
Harnessing metadata can dramatically enhance species identification in bioacoustics, leading to more robust models that adapt to ecological shifts.
Bioacoustic foundation models rely on large-scale citizen science platforms like Xeno-Canto for geographically and ecologically diverse data. Recent work has shown that supervision alone can produce SotA species detection models when trained on this large-scale data -- however, there remains unutilized potential in the form of recording metadata readily available within these community-driven data hubs. In this work, we explore the use of metadata -- such as location and time -- as auxiliary supervision signals, allowing the model to leverage species-metadata correlations in its learned representation. Auxiliary metadata losses provide additional information beyond vocalizations alone that can encourage a richer, more robust representation that generalizes better to species distribution and acoustic domain shifts -- important challenges for deployment in real-world passive acoustic monitoring (PAM) settings. We introduce MetaPerch, a new foundation model that achieves strong species identification performance across multiple challenging domains and present an extensive empirical study of the effects of 9 diverse metadata sources on 17 bioacoustic datasets.