denoising steps

Denoising steps refer to the iterative process employed in certain generative models (like diffusion models) that progressively remove noise from data to produce a clean, coherent output.

12 papers