Generative Modeling with Phase Stochastic Bridges

arXiv:2310.07805v3 Announce Type: replace
Abstract: Diffusion models (DMs) represent state-of-the-art generative models for continuous inputs. DMs work by constructing a Stochastic Differential Equation (SDE) in the input space (ie, position space), and using a neural network to reverse it. In this work, we introduce a novel generative modeling framework grounded in \textbf{phase space dynamics}, where a phase space is defined as {an augmented space encompassing both position and velocity.} Leveraging insights from Stochastic Optimal Control, we construct a path measure …

abstract art arxiv continuous cs.ai cs.lg differential differential equation diffusion diffusion models dynamics equation framework generative generative modeling generative models inputs modeling network neural network novel space state stochastic type work