Speaker Verification on NonverbalTTS NvS 1.0
5.6EERwavlm-base-plus-sv
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| wavlm-base-plus-svSetting=Zero-shot2026.06 | 5.6 | 0.352 | |
| MoE-1 (Data2Vec, 4 experts)Setting=Proposed2026.06 | 9 | 0.527 | |
| (Data2Vec+WavLM) + ECAPASetting=Self-trained2026.06 | 9.24 | 0.528 | |
| MoE-2 (Data2Vec, 4 experts)Setting=Proposed2026.06 | 9.24 | 0.525 | |
| (Voc2Vec+Data2Vec) + ECAPASetting=Self-trained2026.06 | 10.36 | 0.57 | |
| Data2Vec + ECAPA-TDNNSetting=Self-trained2026.06 | 10.76 | 0.611 | |
| WavLM + ECAPA-TDNNSetting=Self-trained2026.06 | 10.92 | 0.571 | |
| (Voc2Vec+WavLM) + ECAPASetting=Self-trained2026.06 | 11.22 | 0.55 | |
| ECAPA-TDNN (Fbank)Setting=Self-trained2026.06 | 11.48 | 0.665 | |
| Voc2Vec + ECAPA-TDNNSetting=Self-trained2026.06 | 11.54 | 0.605 | |
| MoE-2 (Data2Vec, 4 experts)Setting=Proposed2026.06 | 22.66 | 0.817 | |
| Data2Vec + ECAPA-TDNNSetting=Self-trained2026.06 | 23.33 | 0.806 | |
| (Voc2Vec+Data2Vec) + ECAPASetting=Self-trained2026.06 | 23.47 | 0.834 | |
| MoE-1 (Data2Vec, 4 experts)Setting=Proposed2026.06 | 23.95 | 0.833 | |
| (Data2Vec+WavLM) + ECAPASetting=Self-trained2026.06 | 25.46 | 0.848 | |
| Voc2Vec + ECAPA-TDNNSetting=Self-trained2026.06 | 26.16 | 0.881 | |
| (Voc2Vec+WavLM) + ECAPASetting=Self-trained2026.06 | 26.7 | 0.88 | |
| WavLM + ECAPA-TDNNSetting=Self-trained2026.06 | 27.58 | 0.917 | |
| ECAPA-TDNN (Fbank)Setting=Self-trained2026.06 | 28.14 | 0.937 | |
| wavlm-base-plus-svSetting=Zero-shot2026.06 | 38.93 | 0.998 |