Audio Reconstruction on LJSpeech
4.3794UTMOSGT
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GT2025.12 | 4.3794 | — | — | — | — | |
| Q2D2Bandwidth=6.9kbps, Nq=1, token/s=3332025.12 | 4.3302 | 3.4874 | 0.9658 | — | 97.53 | |
| DACBandwidth=9.0kbps, Nq=9, token/s=9002025.12 | 4.3007 | 3.9022 | 0.9733 | — | 97.57 | |
| Q2D2Bandwidth=3.3kbps, Nq=1, token/s=1662025.12 | 4.2909 | 3.2179 | 0.9552 | — | 95.8 | |
| HiFi-CodecBandwidth=3.0kbps, Nq=4, token/s=3002025.12 | 4.2692 | 2.9091 | 0.9485 | — | 94.69 | |
| WavTokenizerBandwidth=0.9kbps, Nq=1, token/s=752025.12 | 4.258 | 2.4923 | 0.9312 | — | 93.97 | |
| SpeechTokenizerBandwidth=6.0kbps, Nq=8, token/s=6002025.12 | 4.2373 | 2.6413 | 0.9316 | — | 94.52 | |
| SpeechTokenizerParams [M]=103.68, Token Rate=50, Codebook Size=1024, Q=8, Bitrate [kbps]=42025.12 | 4.2371 | 2.6411 | 0.9346 | 0.9453 | — | |
| HiFi-CodecBandwidth=4.0kbps, Nq=4, token/s=4002025.12 | 4.1656 | 2.7629 | 0.9446 | — | 94.97 | |
| DACParams [M]=74.71, Token Rate=75, Codebook Size=1024, Q=8, Bitrate [kbps]=62025.12 | 4.0415 | 3.4294 | 0.9567 | 0.967 | — | |
| VocosBandwidth=6.0kbps, Nq=8, token/s=6002025.12 | 4.0332 | 2.9258 | 0.9497 | — | 94.59 | |
| WavTokenizerBandwidth=0.5kbps, Nq=1, token/s=402025.12 | 4.0186 | 2.1142 | 0.9093 | — | 94.06 | |
| SACodecParams [M]=75.17, Token Rate=75, Codebook Size=1000 / 1024, Q=2, Bitrate [kbps]=1.52025.12 | 3.9912 | 2.4249 | 0.9224 | 0.9302 | — | |
| SpeechTokenizerBandwidth=3.0kbps, Nq=4, token/s=3002025.12 | 3.9908 | 2.0458 | 0.9021 | — | 92.99 | |
| FACodecParams [M]=374.49, Token Rate=80, Codebook Size=1024, Q=6, Bitrate [kbps]=4.82025.12 | 3.976 | 2.3234 | 0.922 | 0.9338 | — | |
| Q2D2Bandwidth=1kbps, Nq=1, token/s=752025.12 | 3.9715 | 2.1914 | 0.9158 | — | 92.31 | |
| WavTokenizerParams [M]=80.55, Token Rate=75, Codebook Size=4096, Q=1, Bitrate [kbps]=0.92025.12 | 3.8755 | 1.9532 | 0.9007 | 0.9106 | — | |
| DACBandwidth=4.0kbps, Nq=4, token/s=4002025.12 | 3.8109 | 2.7616 | 0.9338 | — | 95.24 | |
| VocosBandwidth=3.0kbps, Nq=4, token/s=3002025.12 | 3.788 | 2.5006 | 0.931 | — | 93.88 | |
| SpeechTokenizerParams [M]=103.68, Token Rate=50, Codebook Size=1024, Q=3, Bitrate [kbps]=1.52025.12 | 3.5281 | 1.6965 | 0.879 | 0.9154 | — | |
| EncodecBandwidth=6.0kbps, Nq=8, token/s=6002025.12 | 3.2286 | 2.6633 | 0.9441 | — | 95.55 | |
| EncodecParams [M]=14.85, Token Rate=75, Codebook Size=1024, Q=8, Bitrate [kbps]=62025.12 | 3.2281 | 2.6636 | 0.9442 | 0.9555 | — | |
| EncodecBandwidth=3.0kbps, Nq=4, token/s=3002025.12 | 2.3905 | 2.0194 | 0.9058 | — | 93.26 | |
| EncodecParams [M]=14.85, Token Rate=75, Codebook Size=1024, Q=2, Bitrate [kbps]=1.52025.12 | 2.39 | 2.0195 | 0.9058 | 0.9326 | — | |
| DACParams [M]=74.71, Token Rate=75, Codebook Size=1024, Q=2, Bitrate [kbps]=1.52025.12 | 1.8169 | 1.4307 | 0.8487 | 0.8904 | — | |
| DACBandwidth=1.0kbps, Nq=1, token/s=1002025.12 | 1.4438 | 1.2084 | 0.7822 | — | 80.95 |