Audio Generation with Diffusion

Audio Encoder

Module, often based on a VQ-VAE or autoencoder, that compresses a raw audio waveform into a lower-dimensional latent representation, better suited for processing by the diffusion process.

← Tillbaka