Speech Watermarking on LJSpeech (in-distribution)
0.9986Gaussian Noise (5 dB) ScoreVocBulwark (DiffWave)
Evaluation results
| Method | Links | |||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| VocBulwark (DiffWave)backbone=DiffWave, bps=1002026.01 | 0.9986 | 0.9986 | 0.9993 | 0.9993 | 0.9998 | 0.8317 | 0.9889 | 0.9893 | 0.9996 | 0.9977 | 1 | 0.9994 | 0.9992 | 0.9864 | 0.9839 | — | — | — | — | — | — | — | — | |
| Grootbps=1002026.01 | 0.9913 | 0.9939 | 0.9965 | 0.9937 | 0.9866 | 0.9939 | 0.7633 | 0.9961 | 0.9867 | 0.9956 | 1 | — | — | — | — | — | — | — | — | — | — | — | — | |
| VocBulwark (PriorGrad)backbone=PriorGrad, bps=1002026.01 | 0.9172 | 0.9923 | 0.9988 | 0.9996 | 0.9998 | 0.782 | 0.9731 | 0.9431 | 0.9979 | 0.9993 | 1 | 0.9977 | 0.9983 | 0.9863 | 0.9837 | — | — | — | — | — | — | — | — | |
| VocBulwark (HiFi-GAN)backbone=HiFi-GAN, bps=1002026.01 | 0.9076 | 0.9856 | 0.9916 | 0.9965 | 0.9975 | 0.9029 | 0.8749 | 0.9764 | 0.995 | 0.9989 | 1 | 0.991 | 0.8621 | 0.9899 | 0.9923 | — | — | — | — | — | — | — | — | |
| VocBulwark (BigVGAN)backbone=BigVGAN, bps=1002026.01 | 0.8932 | 0.9873 | 0.9968 | 0.999 | 0.9994 | 0.9463 | 0.9939 | 0.8915 | 0.9963 | 0.9998 | 1 | 0.9362 | 0.9479 | 0.9891 | 0.9903 | — | — | — | — | — | — | — | — | |
| RIWFbps=162026.01 | 0.769 | 0.8752 | 0.9508 | 0.9791 | 0.9871 | 0.9255 | 0.6984 | 0.5925 | 0.9462 | 0.9858 | 1 | 0.586 | 0.5872 | 0.8675 | 0.6388 | — | — | — | — | — | — | — | — | |
| FSVCbps=322026.01 | 0.6656 | 0.734 | 0.8116 | 0.8824 | 0.9035 | 0.9873 | 0.8045 | 0.8741 | 0.7872 | 1 | 1 | 0.7382 | 0.7217 | 0.5939 | 0.5842 | — | — | — | — | — | — | — | — | |
| HiFi-GANwbps=202026.01 | 0.6176 | 0.8013 | 0.926 | 0.9702 | 0.9144 | 0.9743 | 0.9827 | 0.9801 | 0.9825 | 0.9893 | 1 | 0.4348 | 0.5254 | 0.9729 | 0.9807 | — | — | — | — | — | — | — | — | |
| AudioSealbps=162026.01 | 0.5951 | 0.6866 | 0.8082 | 0.9104 | 0.9742 | 0.9998 | 0.8852 | 0.8965 | 0.9629 | 0.9998 | 1 | 0.5647 | 0.5063 | 0.9932 | 0.9998 | — | — | — | — | — | — | — | — | |
| Normspacebps=322026.01 | 0.564 | 0.5924 | 0.6121 | 0.6215 | 0.5803 | 0.5308 | 0.4742 | 0.5288 | 0.5708 | 1 | 1 | 0.6542 | 0.6896 | 0.4976 | 0.4949 | — | — | — | — | — | — | — | — | |
| TBWMbps=1002026.01 | 0.56 | 0.6333 | 0.7215 | 0.818 | 0.9938 | 0.9897 | 0.9983 | 0.882 | 0.945 | 0.9999 | 1 | 0.502 | 0.4989 | 0.9988 | 0.9986 | — | — | — | — | — | — | — | — | |
| PBMLbps=1002026.01 | 0.5576 | 0.6101 | 0.6665 | 0.717 | 0.6741 | 0.9743 | 0.7506 | 0.7771 | 0.6993 | 0.9821 | 1 | 0.5057 | 0.5096 | 0.5067 | 0.507 | — | — | — | — | — | — | — | — | |
| WavMarkbps=322026.01 | 0.5056 | 0.5181 | 0.5616 | 0.6513 | 0.9999 | 0.9995 | 1 | 0.9043 | 0.8575 | 1 | 1 | — | — | — | — | — | — | — | — | — | — | — | — | |
| AudioSealbps=162026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.7495 | 0.9743 | 0.576 | 0.7275 | 0.7376 | 0.9342 | 0.5116 | 0.4504 | |
| FSVCbps=322026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.77 | 0.9018 | 0.7605 | 0.7718 | 0.8274 | 0.6699 | 0.5946 | 0.561 | |
| Grootbps=1002026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.9854 | 0.9874 | 0.9917 | 0.9878 | 0.9898 | 0.9809 | 0.7623 | 0.752 | |
| HiFi-GANwbps=202026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.9247 | 0.9146 | 0.8689 | 0.8289 | 0.9508 | 0.4778 | 0.9208 | 0.5235 | |
| Normspacebps=322026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.497 | 0.5804 | 0.5754 | 0.5664 | 0.5061 | 0.6016 | 0.4982 | 0.4978 | |
| PBMLbps=1002026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.6329 | 0.6663 | 0.6408 | 0.594 | 0.6254 | 0.5128 | 0.6085 | 0.5046 | |
| RIWFbps=162026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.7706 | 0.9862 | 0.576 | 0.8905 | 0.8323 | 0.8428 | 0.6187 | 0.607 | |
| TBWMbps=1002026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.6682 | 0.9933 | 0.661 | 0.683 | 0.7541 | 0.5686 | 0.866 | 0.627 | |
| VocBulwarkbps=100, backbone=DiffWave2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.996 | 0.9974 | 0.8656 | 0.9977 | 0.9934 | 0.9973 | 0.99 | 0.9648 | |
| VocBulwarkbps=100, backbone=PriorGrad2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.996 | 0.9974 | 0.8657 | 0.9977 | 0.9921 | 0.9964 | 0.9917 | 0.9605 | |
| VocBulwarkbps=100, backbone=HiFi-GAN2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.9759 | 0.9989 | 0.9471 | 0.9866 | 0.9893 | 0.9725 | 0.9934 | 0.9827 | |
| VocBulwarkbps=100, backbone=BigVGAN2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.9495 | 0.9987 | 0.9029 | 0.9965 | 0.9567 | 0.9851 | 0.9958 | 0.9899 | |
| WavMarkbps=322026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.5059 | 0.9999 | 0.5372 | 0.5443 | 0.7128 | 0.4975 | 0.9345 | 0.4858 |