SAFE Audio Challenge Task 3
SAFE Audio Challenge Task 3 evaluates generated audio detection systems on the third public SAFE challenge task split.
11rows
balanced_accuracyprimary metric
2026-05-06sampled
Metadata
Metrics
generated_g_02, generated_g_04, generated_g_05, generated_g_06, generated_g_09, generated_g_10, generated_g_11, pristine_p_00, pristine_p_01, pristine_p_02, pristine_p_05, pristine_p_09, pristine_p_10, pristine_p_11, pristine_p_16, pristine_p_18, pristine_p_20, generated_accuracy, pristine_accuracy, balanced_accuracy, aug_a_00, aug_a_01, aug_a_02, aug_a_03, aug_a_04, total_time (lower is better), fail_rate (lower is better)
| Rank | Subject | balanced_accuracy | Model Match | Provenance | Sampled |
|---|---|---|---|---|---|
| 1 | baseline-1 | 0.69 | — | Imported | 2026-05-06 |
| 2 | JAIST-HIS | 0.66 | — | Imported | 2026-05-06 |
| 3 | ISSF | 0.66 | — | Imported | 2026-05-06 |
| 4 | ISPL | 0.64 | — | Imported | 2026-05-06 |
| 5 | baseline-2 | 0.62 | — | Imported | 2026-05-06 |
| 6 | Anon_Peking | 0.61 | — | Imported | 2026-05-06 |
| 7 | viper-purdue | 0.58 | — | Imported | 2026-05-06 |
| 8 | gylin | 0.51 | — | Imported | 2026-05-06 |
| 9 | safe-test | 0.51 | — | Imported | 2026-05-06 |
| 10 | DMF | 0.48 | — | Imported | 2026-05-06 |
| 11 | csun22 | 0.47 | — | Imported | 2026-05-06 |
No matching rows.