NVVSpeech-Challenge/NVVSpeech-Challenge-Track2-Test-Set

NVVSpeech Challenge Track 2 Test Set

Track 2 test set for the NVVSpeech Challenge at ISCSLP 2026.

Task

Given a transcript containing one or more non-verbal vocalization (NVV) tags, synthesize speech that naturally realizes the requested NVVs while preserving intelligibility, naturalness, and audio quality.

Dataset Summary

Language	Samples
Chinese	800
English	800
Total	1,600

Files

.
├── README.md
├── SUBMISSION_GUIDE.txt
└── test_set.jsonl

Each line in test_set.jsonl has the following form:

{"id": "en_0001", "text": "Just... [burp] give me a hand with this, will you?"}

• id: unique sample identifier; it must not be changed.
• text: transcript containing the NVV tags to be synthesized.

Canonical NVV Tags

breath, sniff, laugh, cry, cough, throat clearing, sneeze, sigh, gasp, snore, yawn, hum, moan, hiss, lipsmack, and burp.

Citation

The test set is derived from NVV-SuperBench paper. If you use the test set or its evaluation protocol in your work, please cite:

@article{xue2026nvv,
  title={NVV-SuperBench: Beyond Words, Beyond Quality-Benchmarking Nonverbal Vocalizations in Speech Generation},
  author={Xue, Liumeng and Bian, Weizhen and Pan, Jiahao and Wu, Wenxuan and Ren, Yilin and Kang, Boyi and Hu, Jingbin and Ma, Ziyang and Wang, Shuai and Qian, Xinyuan and others},
  journal={arXiv preprint arXiv:2604.16211},
  year={2026}
}