# One-Class Learning with Adaptive Centroid Shift for Audio Deepfake Detection

arXiv: https://arxiv.org/abs/2406.16716 · HTML: https://mlsanity.com/paper/2406.16716/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Jun 2024.

## Built on by (most cited first, 6 of 6)

- [Passive Deepfake Detection Across Multi-modalities: A Comprehensive Survey](https://mlsanity.com/paper/2411.17911/) (2024)
- [Anomaly Detection and Localization for Speech Deepfakes via Feature Pyramid Matching](https://mlsanity.com/paper/2503.18032/) (2025)
- [AFSS: Artifact-Focused Self-Synthesis for Mitigating Bias in Audio Deepfake Detection](https://mlsanity.com/paper/2603.26856/) (2026)
- [End-to-end Audio Deepfake Detection from RAW Waveforms: a RawNet-Based Approach with Cross-Dataset Evaluation](https://mlsanity.com/paper/2504.20923/) (2025)
- [QAMO: Quality-aware Multi-centroid One-class Learning For Speech Deepfake Detection](https://mlsanity.com/paper/2509.20679/) (2025)
- [Learning as Deepfakes Evolve: RF-Prompt for Continual Audio Deepfake Detection](https://mlsanity.com/paper/2609.37586/) (2026)

## Abstract (arXiv)

As speech synthesis systems continue to make remarkable advances in recent years, the importance of robust deepfake detection systems that perform well in unseen systems has grown. In this paper, we propose a novel adaptive centroid shift (ACS) method that updates the centroid representation by continually shifting as the weighted average of bonafide representations. Our approach uses only bonafide samples to define their centroid, which can yield a specialized centroid for one-class learning. Integrating our ACS with one-class learning gathers bonafide representations into a single cluster, forming well-separated embeddings robust to unseen spoofing attacks. Our proposed method achieves an equal error rate (EER) of 2.19% on the ASVspoof 2021 deepfake dataset, outperforming all existing systems. Furthermore, the t-SNE visualization illustrates that our method effectively maps the bonafide embeddings into a single cluster and successfully disentangles the bonafide and spoof classes.