Target-Speaker Multilingual Speech Generation on TTS multilingual Korean (test)
1.741WERQwen3-TTS
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-TTSTokenizer Frequency=12Hz, Parameters=1.7B, Evaluation Protocol=CustomVoice2026.01 | 1.741 | |
| Qwen3-TTSTokenizer Frequency=25Hz, Parameters=0.6B, Evaluation Protocol=CustomVoice2026.01 | 2.202 | |
| Qwen3-TTSTokenizer Frequency=25Hz, Parameters=1.7B, Evaluation Protocol=CustomVoice2026.01 | 2.274 | |
| GPT-4o-Audio Preview2026.01 | 2.763 | |
| Qwen3-TTSTokenizer Frequency=12Hz, Parameters=0.6B, Evaluation Protocol=CustomVoice2026.01 | 3.053 |