Temporal Pooling Strategies for Training-Free Anomalous Sound Detection with Self-Supervised Audio Embeddings

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wilkinghoff, Kevin, Yadav, Sarthak, Tan, Zheng-Hua
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917315235282944
author Wilkinghoff, Kevin
Yadav, Sarthak
Tan, Zheng-Hua
author_facet Wilkinghoff, Kevin
Yadav, Sarthak
Tan, Zheng-Hua
contents Training-free anomalous sound detection (ASD) based on pre-trained audio embedding models has recently garnered significant attention, as it enables the detection of anomalous sounds using only normal reference data while offering improved robustness under domain shifts. However, existing embedding-based approaches almost exclusively rely on temporal mean pooling, while alternative pooling strategies have so far only been explored for spectrogram-based representations. Consequently, the role of temporal pooling in training-free ASD with pre-trained embeddings remains insufficiently understood. In this paper, we present a systematic evaluation of temporal pooling strategies across multiple state-of-the-art audio embedding models. We propose relative deviation pooling (RDP), an adaptive pooling method that emphasizes informative temporal deviations, and introduce a hybrid pooling strategy that combines RDP with generalized mean pooling. Experiments on five benchmark datasets demonstrate that the proposed methods consistently outperform mean pooling and achieve state-of-the-art performance for training-free ASD, including results that surpass all previously reported trained systems and ensembles on the DCASE2025 ASD dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2603_04605
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Temporal Pooling Strategies for Training-Free Anomalous Sound Detection with Self-Supervised Audio Embeddings
Wilkinghoff, Kevin
Yadav, Sarthak
Tan, Zheng-Hua
Audio and Speech Processing
Sound
Training-free anomalous sound detection (ASD) based on pre-trained audio embedding models has recently garnered significant attention, as it enables the detection of anomalous sounds using only normal reference data while offering improved robustness under domain shifts. However, existing embedding-based approaches almost exclusively rely on temporal mean pooling, while alternative pooling strategies have so far only been explored for spectrogram-based representations. Consequently, the role of temporal pooling in training-free ASD with pre-trained embeddings remains insufficiently understood. In this paper, we present a systematic evaluation of temporal pooling strategies across multiple state-of-the-art audio embedding models. We propose relative deviation pooling (RDP), an adaptive pooling method that emphasizes informative temporal deviations, and introduce a hybrid pooling strategy that combines RDP with generalized mean pooling. Experiments on five benchmark datasets demonstrate that the proposed methods consistently outperform mean pooling and achieve state-of-the-art performance for training-free ASD, including results that surpass all previously reported trained systems and ensembles on the DCASE2025 ASD dataset.
title Temporal Pooling Strategies for Training-Free Anomalous Sound Detection with Self-Supervised Audio Embeddings
topic Audio and Speech Processing
Sound
url https://arxiv.org/abs/2603.04605