From Sharpness to Better Generalization for Speech Deepfake Detection
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911004605022208 |
|---|---|
| author | Huang, Wen Liu, Xuechen Wang, Xin Yamagishi, Junichi Qian, Yanmin |
| author_facet | Huang, Wen Liu, Xuechen Wang, Xin Yamagishi, Junichi Qian, Yanmin |
| contents | Generalization remains a critical challenge in speech deepfake detection (SDD). While various approaches aim to improve robustness, generalization is typically assessed through performance metrics like equal error rate without a theoretical framework to explain model performance. This work investigates sharpness as a theoretical proxy for generalization in SDD. We analyze how sharpness responds to domain shifts and find it increases in unseen conditions, indicating higher model sensitivity. Based on this, we apply Sharpness-Aware Minimization (SAM) to reduce sharpness explicitly, leading to better and more stable performance across diverse unseen test sets. Furthermore, correlation analysis confirms a statistically significant relationship between sharpness and generalization in most test settings. These findings suggest that sharpness can serve as a theoretical indicator for generalization in SDD and that sharpness-aware training offers a promising strategy for improving robustness. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_11532 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | From Sharpness to Better Generalization for Speech Deepfake Detection Huang, Wen Liu, Xuechen Wang, Xin Yamagishi, Junichi Qian, Yanmin Audio and Speech Processing Sound Generalization remains a critical challenge in speech deepfake detection (SDD). While various approaches aim to improve robustness, generalization is typically assessed through performance metrics like equal error rate without a theoretical framework to explain model performance. This work investigates sharpness as a theoretical proxy for generalization in SDD. We analyze how sharpness responds to domain shifts and find it increases in unseen conditions, indicating higher model sensitivity. Based on this, we apply Sharpness-Aware Minimization (SAM) to reduce sharpness explicitly, leading to better and more stable performance across diverse unseen test sets. Furthermore, correlation analysis confirms a statistically significant relationship between sharpness and generalization in most test settings. These findings suggest that sharpness can serve as a theoretical indicator for generalization in SDD and that sharpness-aware training offers a promising strategy for improving robustness. |
| title | From Sharpness to Better Generalization for Speech Deepfake Detection |
| topic | Audio and Speech Processing Sound |
| url | https://arxiv.org/abs/2506.11532 |