NVBench: A Benchmark for Speech Synthesis with Non-Verbal Vocalizations
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915946878205952 |
|---|---|
| author | Xue, Liumeng Bian, Weizhen Pan, Jiahao Wang, Wenxuan Ren, Yilin Kang, Boyi Hu, Jingbin Ma, Ziyang Wang, Shuai Qian, Xinyuan Lee, Hung-yi Guo, Yike |
| author_facet | Xue, Liumeng Bian, Weizhen Pan, Jiahao Wang, Wenxuan Ren, Yilin Kang, Boyi Hu, Jingbin Ma, Ziyang Wang, Shuai Qian, Xinyuan Lee, Hung-yi Guo, Yike |
| contents | Non-verbal vocalizations (NVVs) like laugh, sigh, and sob are essential for human-like speech, yet standardized evaluation remains limited in jointly assessing whether systems can generate the intended NVVs, place them correctly, and keep them salient without harming speech. We present Non-verbal Vocalization Benchmark (NVBench), a bilingual (English/Chinese) benchmark that evaluates speech synthesis with NVVs. NVBench pairs a unified 45-type taxonomy with a curated bilingual dataset and introduces a multi-axis protocol that separates general speech naturalness and quality from NVV-specific controllability, placement, and salience. We benchmark 15 TTS systems using objective metrics, listening tests, and an LLM-based multi-rater evaluation. Results reveal that NVVs controllability often decouples from quality, while low-SNR oral cues and long-duration affective NVVs remain persistent bottlenecks. NVBench enables fair cross-system comparison across diverse control interfaces under a unified, standardized framework. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_16211 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | NVBench: A Benchmark for Speech Synthesis with Non-Verbal Vocalizations Xue, Liumeng Bian, Weizhen Pan, Jiahao Wang, Wenxuan Ren, Yilin Kang, Boyi Hu, Jingbin Ma, Ziyang Wang, Shuai Qian, Xinyuan Lee, Hung-yi Guo, Yike Sound Non-verbal vocalizations (NVVs) like laugh, sigh, and sob are essential for human-like speech, yet standardized evaluation remains limited in jointly assessing whether systems can generate the intended NVVs, place them correctly, and keep them salient without harming speech. We present Non-verbal Vocalization Benchmark (NVBench), a bilingual (English/Chinese) benchmark that evaluates speech synthesis with NVVs. NVBench pairs a unified 45-type taxonomy with a curated bilingual dataset and introduces a multi-axis protocol that separates general speech naturalness and quality from NVV-specific controllability, placement, and salience. We benchmark 15 TTS systems using objective metrics, listening tests, and an LLM-based multi-rater evaluation. Results reveal that NVVs controllability often decouples from quality, while low-SNR oral cues and long-duration affective NVVs remain persistent bottlenecks. NVBench enables fair cross-system comparison across diverse control interfaces under a unified, standardized framework. |
| title | NVBench: A Benchmark for Speech Synthesis with Non-Verbal Vocalizations |
| topic | Sound |
| url | https://arxiv.org/abs/2604.16211 |