Risk-Sensitive Diffusion: Robustly Optimizing Diffusion Models with Noisy Samples

Understand

Diffusion models are mainly studied on image data.

• However, non-image data (e.g., tabular data) are also prevalent in real applications and tend to be noisy due to some inevitable factors in the stage of data collection, degrading the generation quality of diffusion models.
• In this paper, we consider a novel problem setting where every collected sample is paired with a vector indicating the data quality: risk vector.
• This setting applies to many scenarios involving noisy data and we propose risk-sensitive SDE, a type of stochastic differential equation (SDE) parameterized by the risk vector, to address it.