EUVIMEDEUVIMED entdecken
Crossref

Sidon: High fidelity robust multilingual speech restoration based on parametric speech resynthesis

Wataru Nakata, Yuki Saito, Hiroshi Saruwatari

The Journal of the Acoustical Society of America · 01.10.2025 · 158 (4_Supplement) · A261-A261

DOI: 10.1121/10.0040823 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of the Acoustical Society of America, erschienen am 01.10.2025. Verfasst von Wataru Nakata und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Speech restoration is becoming a foundational technology as it enables to use in-the-wild speech data for the generative AI training by removing artifacts. We present Sidon: a robust and high fidelity speech restoration model based on parametric speech resynthesis. Sidon is highly multilingual with 102 languages and outputs full fidelity speech in 48 kHz sampling rate. Upon training multiple artifacts are simulated such as background noise, reverberation, band limiting, codec, clipping, wind noise and packet loss. We evaluate the performance of Sidon against Google's internal Miipher and Miipher-2. The result shows competitive performance in several metrics. We also show that the Sidon is well generalized to the in-the-wild noisy speech recordings.

Zitieren

APA

Nakata, W., Saito, Y., & Saruwatari, H. (2025). Sidon: High fidelity robust multilingual speech restoration based on parametric speech resynthesis. The Journal of the Acoustical Society of America, 158(4_Supplement), A261-A261. https://doi.org/10.1121/10.0040823

Vancouver

Nakata W, Saito Y, Saruwatari H. Sidon: High fidelity robust multilingual speech restoration based on parametric speech resynthesis. The Journal of the Acoustical Society of America. 2025;158(4_Supplement):A261-A261. doi:10.1121/10.0040823

BibTeX
@article{Nakata2025,
  author = {Nakata, Wataru and Saito, Yuki and Saruwatari, Hiroshi},
  title = {Sidon: High fidelity robust multilingual speech restoration based on parametric speech resynthesis},
  journal = {The Journal of the Acoustical Society of America},
  year = {2025},
  volume = {158},
  number = {4_Supplement},
  pages = {A261--A261},
  doi = {10.1121/10.0040823}
}
Kostenlos starten →