Training Data Protection with Compositional Diffusion Models. (arXiv:2308.01937v1 [cs.LG])

We introduce Compartmentalized Diffusion Models (CDM), a method to train
different diffusion models (or prompts) on distinct data sources and
arbitrarily compose them at inference time. The individual models can be
trained in isolation, at different times, and on different distributions and
domains and can be later composed to achieve performance comparable to a
paragon model trained on all data simultaneously. Furthermore, each model only
contains information about the subset of the data it was exposed to during
training, enabling …

cdm compose data data protection data sources diffusion models distributions domains isolation performance prompts protection train training