MelGAN: convolutional adversarial waveform synthesis

What changed

Developed a non-autoregressive convolutional generator and multi-scale adversarial training choices for conditional mel-spectrogram inversion.

Scientific scope

Developed a non-autoregressive convolutional generator and multi-scale adversarial training choices for conditional mel-spectrogram inversion. Inherits GANs and convolutional waveform modeling. Acquired delta is architecture/training integration and ablation, not first adversarial audio. Unseen-speaker examples and speed comparisons do not establish universal quality/generalization.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits GANs and convolutional waveform modeling. Acquired delta is architecture/training integration and ablation, not first adversarial audio. Unseen-speaker examples and speed comparisons do not establish universal quality/generalization.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• MelGAN: Generative Adversarial Networks for Conditional Waveform Synthesis

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction