Metraipolis › archive › dc_daddf9ff1d65
ASR-Roundtrip Evaluation Can Mask Context- and Convention-Dependent Reading Errors in Chinese News TTS — arXiv:2608.10606
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.10606
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T06:14:12.157Z
- SHA-256 of the archived text
62f5fef74dc948b0463f48de108e2c6d71a2106c693e1aefcbbaaa24f0dd7fd6- Stable URL of this page
https://metraipolis.com/documento/dc_daddf9ff1d65
How to cite
ASR-Roundtrip Evaluation Can Mask Context- and Convention-Dependent Reading Errors in Chinese News TTS — arXiv:2608.10606 — arXiv (en), https://arxiv.org/abs/2608.10606 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_daddf9ff1d65 (Metraipolis archive dc_daddf9ff1d65, sha256 62f5fef74dc948b0463f48de108e2c6d71a2106c693e1aefcbbaaa24f0dd7fd6)
BibTeX
@misc{dcdaddf9ff1d65,
title = {ASR-Roundtrip Evaluation Can Mask Context- and Convention-Dependent Reading Errors in Chinese News TTS — arXiv:2608.10606},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.10606},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_daddf9ff1d65; sha256 62f5fef74dc948b0463f48de108e2c6d71a2106c693e1aefcbbaaa24f0dd7fd6},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.10606. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
ASR-roundtrip evaluation is widely used as a scalable proxy for text-to-speech (TTS) intelligibility, but it can produce false negatives for reading errors perceived by listeners. We study Chinese news TTS spans whose correct reading depends on context or domain conventions, such as sports scores, aircraft models, technical units, and membership names. In these cases, Raw TTS can choose a plausible but wrong reading while ASR transcribes the audio as the intended or surface-correct text. A targeted audit over 110 high-risk MiMo TTS cases, reported with a complete denominator, confirms 46 masked false negatives, 9 exposed TTS errors, and 55 cases with no Raw TTS error. A span-isolation diagnostic re-exposes 18/46 previously masked errors. A Raw-only CosyVoice audit on the same targeted pool confirms 51 masked cases. Across the 97 TTS-specific audio files labeled confirmed masked across the two audits, Qwen3-ASR surface-recovers 40 cases, whereas Paraformer does so in only 2. The results suggest that ASR-roundtrip is useful for screening but insufficient as standalone ground truth for Chinese news reading-risk evaluation. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-12 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.10606 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-19T06:08:50.318Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).