Table 5 BLSTM accuracy as a function of the hidden layer size
From: Advanced recurrent network-based hybrid acoustic models for low resource speech recognition
Model | # hidden layer size | Â | |||
|---|---|---|---|---|---|
| Â | 400 | 500 | 600 | 700 | 800 |
BLSTM-fbank | 48.6 | 48.3 | 48.7 | 49.2 | 49.5 |
BLSTM-MBN | 44 | 43.8 | 43.9 | 44.5 | 44.8 |