Table 7 Comparisons on different heads for single attention
From: A new joint CTC-attention-based speech recognition model with multi-level multi-head attention
| Â | 2 heads | 3 heads | 4 heads | 5 heads |
|---|---|---|---|---|
TIMIT | 16.84 | 16.65 | 16.51 | 16.78 |
WSJ | 4.5 | 4.2 | 4.1 | 4.4 |
LibriSpeech | 4.2 | 4.0 | 3.8 | 4.1 |