Table 3 LSD of the signals reproduced by different BWE methods
From: Audio bandwidth extension using ensemble of recurrent neural networks
Data type | HMM | EESN |
|---|---|---|
Country music | 6.6788 | 7.5653 |
Jazz music | 12.6068 | 8.4862 |
Rock music | 6.3262 | 6.4720 |
Violin solo | 3.8470 | 3.5376 |
Symphony | 3.9713 | 5.2903 |
Speech | 6.1027 | 5.8650 |
Average | 6.5888 | 6.2027 |