A Causal Bootstrap for Diverse Image Generation under Unobserved Confounding

The paper introduces CauVaDE, a causal generative model that represents an unobserved confounder as a discrete latent cluster, enabling accurate approximation of both observational and interventional image distributions. It demonstrates that this representation can capture diverse causal explanations without sacrificing fit to the data. A novel causal bootstrap method is proposed to vary confounder entropy, producing a family of distinct interventional image distributions.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

• rag
• diffusion