# No Free Lunch from Audio Pretraining in Bioacoustics: A Benchmark Study of Embeddings

arXiv: https://arxiv.org/abs/2508.10230 · HTML: https://mlsanity.com/paper/2508.10230/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 30 references that are in the map.
- First posted on arXiv 13 Aug 2025.

## Builds on (most cited first, 10 of 10)

- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [CNN Architectures for Large-Scale Audio Classification](https://mlsanity.com/paper/1609.09430/) (2016)
- [Delving into Deep Imbalanced Regression](https://mlsanity.com/paper/2102.09554/) (2021)
- [Computational bioacoustics with deep learning: a review and roadmap](https://mlsanity.com/paper/2112.06725/) (2021)
- [AVES: Animal Vocalization Encoder based on Self-Supervision](https://mlsanity.com/paper/2210.14493/) (2022)
- [Global birdsong embeddings enable superior transfer learning for bioacoustic classification](https://mlsanity.com/paper/2307.06292/) (2023)
- [BEANS: The Benchmark of Animal Sounds](https://mlsanity.com/paper/2210.12300/) (2022)
- [Data augmentation approaches for improving animal audio classification](https://mlsanity.com/paper/1912.07756/) (2019)
- [Can Self-Supervised Neural Representations Pre-Trained on Human Speech distinguish Animal Callers?](https://mlsanity.com/paper/2305.14035/) (2023)
- [Rank-based loss for learning hierarchical representations](https://mlsanity.com/paper/2110.05941/) (2021)

## Built on by (most cited first, 1 of 1)

- [DeepForestSound: a multi-species automatic detector for passive acoustic monitoring in African tropical forests, a case study in Kibale National Park](https://mlsanity.com/paper/2604.08087/) (2026)

## Abstract (arXiv)

Bioacoustics, the study of animal sounds, offers a non-invasive method to monitor ecosystems. Extracting embeddings from audio-pretrained deep learning (DL) models without fine-tuning has become popular for obtaining bioacoustic features for tasks. However, a recent benchmark study reveals that while fine-tuned audio-pretrained VGG and transformer models achieve state-of-the-art performance in some tasks, they fail in others. This study benchmarks 11 DL models on the same tasks by reducing their learned embeddings' dimensionality and evaluating them through clustering. We found that audio-pretrained DL models 1) without fine-tuning even underperform fine-tuned AlexNet, 2) both with and without fine-tuning fail to separate the background from labeled sounds, but ResNet does, and 3) outperform other models when fewer background sounds are included during fine-tuning. This study underscores the necessity of fine-tuning audio-pretrained models and checking the embeddings after fine-tuning. Our codes are available: https://github.com/NeuroscienceAI/AudioEmbeddings