Synthetic data enables context-aware bioacoustic sound event detection
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911139780100096 |
|---|---|
| author | Hoffman, Benjamin Robinson, David Miron, Marius Baglione, Vittorio Canestrari, Daniela Elias, Damian Trapote, Eva Effenberger, Felix Cusimano, Maddie Hagiwara, Masato Pietquin, Olivier |
| author_facet | Hoffman, Benjamin Robinson, David Miron, Marius Baglione, Vittorio Canestrari, Daniela Elias, Damian Trapote, Eva Effenberger, Felix Cusimano, Maddie Hagiwara, Masato Pietquin, Olivier |
| contents | We propose a methodology for training foundation models that enhances their in-context learning capabilities within the domain of bioacoustic signal processing. We use synthetically generated training data, introducing a domain-randomization-based pipeline that constructs diverse acoustic scenes with temporally strong labels. We generate over 8.8 thousand hours of strongly-labeled audio and train a query-by-example, transformer-based model to perform few-shot bioacoustic sound event detection. Our second contribution is a public benchmark of 13 diverse few-shot bioacoustics tasks. Our model outperforms previously published methods, and improves relative to other training-free methods by $64\%$. We demonstrate that this is due to increase in model size and data scale, as well as algorithmic improvements. We make our trained model available via an API, to provide ecologists and ethologists with a training-free tool for bioacoustic sound event detection. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_00296 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Synthetic data enables context-aware bioacoustic sound event detection Hoffman, Benjamin Robinson, David Miron, Marius Baglione, Vittorio Canestrari, Daniela Elias, Damian Trapote, Eva Effenberger, Felix Cusimano, Maddie Hagiwara, Masato Pietquin, Olivier Sound Machine Learning Audio and Speech Processing We propose a methodology for training foundation models that enhances their in-context learning capabilities within the domain of bioacoustic signal processing. We use synthetically generated training data, introducing a domain-randomization-based pipeline that constructs diverse acoustic scenes with temporally strong labels. We generate over 8.8 thousand hours of strongly-labeled audio and train a query-by-example, transformer-based model to perform few-shot bioacoustic sound event detection. Our second contribution is a public benchmark of 13 diverse few-shot bioacoustics tasks. Our model outperforms previously published methods, and improves relative to other training-free methods by $64\%$. We demonstrate that this is due to increase in model size and data scale, as well as algorithmic improvements. We make our trained model available via an API, to provide ecologists and ethologists with a training-free tool for bioacoustic sound event detection. |
| title | Synthetic data enables context-aware bioacoustic sound event detection |
| topic | Sound Machine Learning Audio and Speech Processing |
| url | https://arxiv.org/abs/2503.00296 |