Audio Synthesis on Singing Voice Academic setting
4.18MOS Prediction ScoreGround Truth
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Ground TruthBitrate=/, # Param=/2025.12 | 4.18 | 0 | 0 | |
| BigCodecBitrate=8 kbps, # Param=412M2025.12 | 4.17 | 20.44 | 0.0598 | |
| Pupu-CodeclargeBitrate=8 kbps, # Param=119M2025.12 | 4.17 | 19.84 | 0.0528 | |
| DACBitrate=8 kbps, # Param=154M2025.12 | 4.16 | 19.57 | 0.0557 | |
| DACBitrate=5.33 kbps, # Param=154M2025.12 | 4.15 | 21.5 | 0.063 | |
| Pupu-CodeclargeBitrate=5.33 kbps, # Param=119M2025.12 | 4.15 | 23.53 | 0.0624 | |
| BigCodecBitrate=5.33 kbps, # Param=412M2025.12 | 4.14 | 24.05 | 0.0719 | |
| Pupu-CodecsmallBitrate=8 kbps, # Param=32M2025.12 | 4.1 | 22.31 | 0.0559 | |
| DACBitrate=2.67 kbps, # Param=154M2025.12 | 4.06 | 27.05 | 0.0825 | |
| Pupu-CodeclargeBitrate=2.67 kbps, # Param=119M2025.12 | 4.06 | 29.84 | 0.0841 | |
| Pupu-CodecsmallBitrate=5.33 kbps, # Param=32M2025.12 | 4.05 | 25.52 | 0.0667 | |
| BigCodecBitrate=2.67 kbps, # Param=412M2025.12 | 4.02 | 30.37 | 0.093 | |
| EnCodecBitrate=8 kbps, # Param=59M2025.12 | 3.97 | 26.72 | 0.0875 | |
| DACBitrate=1.78 kbps, # Param=154M2025.12 | 3.94 | 33.34 | 0.1005 | |
| Pupu-CodecsmallBitrate=2.67 kbps, # Param=32M2025.12 | 3.87 | 32.64 | 0.0901 | |
| Pupu-CodeclargeBitrate=1.78 kbps, # Param=119M2025.12 | 3.87 | 35.51 | 0.1018 | |
| BigCodecBitrate=1.78 kbps, # Param=412M2025.12 | 3.84 | 36.35 | 0.1051 | |
| EnCodecBitrate=5.33 kbps, # Param=59M2025.12 | 3.82 | 30.79 | 0.0993 | |
| Pupu-CodecsmallBitrate=1.78 kbps, # Param=32M2025.12 | 3.65 | 39.16 | 0.1133 | |
| EnCodecBitrate=2.67 kbps, # Param=59M2025.12 | 3.34 | 44.99 | 0.141 | |
| EnCodecBitrate=1.78 kbps, # Param=59M2025.12 | 2.64 | 64.44 | 0.1967 |