Denoising
The process of removing noise from data, a fundamental operation in diffusion models where the network learns to progressively remove Gaussian noise to generate clean data.
In Diffusion Models
The denoising process is the generative heart of diffusion models. Starting from pure noise, the model predicts and removes noise at each timestep, gradually revealing a coherent image, audio, or other output.
Denoising Autoencoders
A training technique where the model learns to reconstruct clean data from corrupted inputs. This forces the model to learn robust features rather than memorizing inputs, and is used as a pre-training objective.