Lifting Architectural Constraints of Injective Flows

arXiv:2306.01843v4 Announce Type: replace
Abstract: Normalizing Flows explicitly maximize a full-dimensional likelihood on the training data. However, real data is typically only supported on a lower-dimensional manifold leading the model to expend significant compute on modeling noise. Injective Flows fix this by jointly learning a manifold and the distribution on it. So far, they have been limited by restrictive architectures and/or high computational cost. We lift both constraints by a new efficient estimator for the maximum likelihood loss, compatible with …

abstract arxiv compute constraints cs.lg data distribution however likelihood manifold modeling noise real data training training data type