Augmented Normalizing Flows: Bridging the Gap Between Generative Flows and Latent Variable Models
This work addresses a bottleneck in generative modeling for researchers and practitioners by bridging flows and latent variable models, though it appears incremental.
The authors tackled the problem of improving expressivity in generative flows without significantly increasing computational costs, achieving state-of-the-art performance on standard benchmarks.
In this work, we propose a new family of generative flows on an augmented data space, with an aim to improve expressivity without drastically increasing the computational cost of sampling and evaluation of a lower bound on the likelihood. Theoretically, we prove the proposed flow can approximate a Hamiltonian ODE as a universal transport map. Empirically, we demonstrate state-of-the-art performance on standard benchmarks of flow-based generative modeling.