Audio Reconstruction on Common Voice
0.9888CVUTokenBridge
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| TokenBridgeCodebook Size=40962026.02 | 0.9888 | 1.1017 | 0.2787 | |
| TokenBridgeCodebook Size=2562026.02 | 0.9887 | 1.0967 | 0.2688 | |
| TokenBridgeCodebook Size=10242026.02 | 0.9887 | 1.1059 | 0.2856 | |
| TokenBridgeCodebook Size=163842026.02 | 0.9887 | 1.1202 | 0.2722 | |
| FSPCodebook Size=2562026.02 | 0.286 | 1.8148 | 0.8537 | |
| VP-VAECodebook Size=2562026.02 | 0.2055 | 1.851 | 0.8617 | |
| FSPCodebook Size=10242026.02 | 0.1776 | 1.9549 | 0.8676 | |
| FSQCodebook Size=2562026.02 | 0.169 | 1.2064 | 0.763 | |
| VP-VAECodebook Size=10242026.02 | 0.1441 | 1.9697 | 0.8701 | |
| VP-VAECodebook Size=163842026.02 | 0.1218 | 2.1803 | 0.9003 | |
| VP-VAECodebook Size=40962026.02 | 0.1151 | 2.0841 | 0.8907 | |
| FSQCodebook Size=10242026.02 | 0.1012 | 1.7776 | 0.8498 | |
| FSPCodebook Size=40962026.02 | 0.0936 | 2.0564 | 0.884 | |
| FSQCodebook Size=40962026.02 | 0.067 | 1.9712 | 0.8707 | |
| FSPCodebook Size=163842026.02 | 0.0541 | 2.138 | 0.8906 | |
| FSQCodebook Size=163842026.02 | 0.0499 | 1.9566 | 0.8721 | |
| SimVQCodebook Size=2562026.02 | 0.0041 | 1.2103 | 0.3781 | |
| VQ-VAECodebook Size=10242026.02 | 0.001 | 1.2103 | 0.3371 | |
| SimVQCodebook Size=10242026.02 | 0.001 | 1.1568 | 0.3627 | |
| SimVQCodebook Size=40962026.02 | 0.0002 | 1.144 | 0.3713 | |
| SimVQCodebook Size=163842026.02 | 0.0001 | 1.1807 | 0.4002 |