A Universal Harmonic Discriminator for High-quality GAN-based Vocoder
Nan Xu, Zhaolong Huang, Xiao Zeng* · 2025
Read A Universal Harmonic Discriminator for High-quality GAN-based Vocoder by Nan Xu, Zhaolong Huang, Xiao Zeng* (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-03486.txt.
Description
Synthetic voices need sharp critics to sound real. Nan Xu, Zhaolong Huang, and Xiao Zeng present a universal harmonic discriminator in this 2025 work. Vocoders based on GANs use a discriminator to judge the generated audio.
The team focuses on time-frequency discriminators, which often read a Short-Time Fourier Transform of the sound. Their harmonic design aims to judge voices across generators, hence universal. Better discriminators push vocoders toward natural speech. That helps text-to-speech tools used in apps, books, and accessibility. Harmonic structure matches speech, where pitch and its overtones define voice quality.
Catalog record
| CREATOR | Nan Xu, Zhaolong Huang, Xiao Zeng* |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-03486