Diffusion Models for Audio

Audio U-Net

Encoder-decoder neural network architecture with skip connections, adapted to process spectrograms and predict noise at each step of the audio denoising process.

← Geri