PEER-REVIEWED PAPERBioacoustic Foundation ModelsDOI: 10.1109/ICASSP49357.2023.10095642
AVES: Animal Vocalization Encoder Based on Self-Supervision
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)2023-06-04
Research Abstract & Summary
Peer-reviewed conference paper by Masato Hagiwara introducing AVES, a self-supervised HuBERT-derived transformer pre-trained on diverse animal bioacoustic audio datasets, outperforming general audio encoders on downstream bioacoustic classification tasks.
Scholarly Resolver: Canonical literature access