Table 1 AP17-OLR dataset
From: Masked multi-center angular margin loss for language recognition
Language | Train/dev | Test | ||
|---|---|---|---|---|
Speaker | All utt. | Speaker | All utt. | |
ka-cn | 86 | 4200 | 86 | 1800 |
ti-cn | 34 | 11,100 | 34 | 1800 |
uy-id | 353 | 5163 | 353 | 1800 |
ct-cn | 24 | 7676 | 6 | 2556 |
zh-cn | 24 | 7198 | 6 | 2400 |
id-id | 24 | 7667 | 6 | 2557 |
ja-jp | 24 | 7655 | 6 | 2548 |
ru-ru | 24 | 7183 | 6 | 1800 |
ko-kr | 24 | 7195 | 6 | 2398 |
vi-vn | 24 | 7197 | 6 | 2396 |