SAFE Audio Challenge Task 1: leaderboard
SAFE Audio Challenge Task 1 evaluates systems for detecting generated audio against pristine audio sources.
Metric: Balanced Accuracy (%). Source: huggingface.co. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | ISPL | 89.64 | |
| 2 | baseline-2 | 85.77 | |
| 3 | baseline-1 | 81.89 | |
| 4 | ISSF | 80.99 | |
| 5 | Anon_Peking | 76.92 | |
| 6 | viper-purdue | 76.81 | |
| 7 | JAIST-HIS | 72.91 | |
| 8 | DMF | 64.31 | |
| 9 | ZJM0813 | 64.29 | |
| 10 | gylin | 54.26 |
Interactive version: theaggregate.ai/benchmark?slug=safe-audio-challenge-task-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.