Bemaganv2: Discriminator Combination Strategies For Gan-based Vocoders In Long-term Audio Generation
2025 · Taesoo Park, Mungwi Jeong, Mingyu Park, et al.
Abstract
This paper presents BemaGANv2, an advanced GAN-based vocoder designed for high-fidelity and long-term audio generation, with a focus on systematic evaluation of discriminator combination strategies. Long-term audio generation is critical for applications in Text-to-Music (TTM) and Text-to-Audio (TTA) systems, where maintaining temporal co- herence, prosodic consistency, and harmonic structure over extended durations remains a significant challenge. Built upon the original BemaGAN architecture, BemaGANv2 incorporates major architectural innovations by replacing traditional ResBlocks in the generator with the Anti-aliased Multi-Periodicity composition (AMP) module, which internally applies the Snake activation function to better model periodic structures. In the discriminator framework, we integrate the Multi-Envelope Discriminator (MED), a novel architecture we proposed, to extract rich temporal en- velope features crucial for periodicity detection. Coupled with the Multi-Resolution Dis
Authors
(none)
Tags
Stats
Related papers
- Vnet: A Gan-based Multi-tier Discriminator Network For Speech Synthesis Vocoders (2024)2.26
- Vocgan: A High-fidelity Real-time Vocoder With A Hierarchically-nested Adversarial Network (2020)12.54
- Bigvgan: A Universal Neural Vocoder With Large-scale Training (2022)6.17
- A Multi-scale Time-frequency Spectrogram Discriminator For Gan-based Non-autoregressive TTS (2022)6.77
- Enhancing Gan-based Vocoders With Contrastive Learning Under Data-limited Condition (2023)3.58
- Snakegan: A Universal Vocoder Leveraging DDSP Prior Knowledge And Periodic Inductive Bias (2023)4.52
- TFGAN: Time And Frequency Domain Based Generative Adversarial Network For High-fidelity Speech Synthesis (2020)0.00
- DSPGAN: A Gan-based Universal Vocoder For High-fidelity TTS By Time-frequency Domain Supervision From DSP (2022)9.03