# Improvements of Discriminative Feature Space Training for Anomalous Sound Detection in Unlabeled Conditions

arXiv: https://arxiv.org/abs/2409.09332 · HTML: https://mlsanity.com/paper/2409.09332/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 14 Sep 2024.

## Built on by (most cited first, 2 of 2)

- [Improving Anomalous Sound Detection with Attribute-aware Representation from Domain-adaptive Pre-training](https://mlsanity.com/paper/2509.12845/) (2025)
- [Sub-Band Spectral Matching with Localized Score Aggregation for Robust Anomalous Sound Detection](https://mlsanity.com/paper/2603.13749/) (2026)

## Abstract (arXiv)

In anomalous sound detection, the discriminative method has demonstrated superior performance. This approach constructs a discriminative feature space through the classification of the meta-information labels for normal sounds. This feature space reflects the differences in machine sounds and effectively captures anomalous sounds. However, its performance significantly degrades when the meta-information labels are missing. In this paper, we improve the performance of a discriminative method under unlabeled conditions by two approaches. First, we enhance the feature extractor to perform better under unlabeled conditions. Our enhanced feature extractor utilizes multi-resolution spectrograms with a new training strategy. Second, we propose various pseudo-labeling methods to effectively train the feature extractor. The experimental evaluations show that the proposed feature extractor and pseudo-labeling methods significantly improve performance under unlabeled conditions.