Multi-Resolution Continuous Normalizing Flows
This work addresses a scalability challenge in generative modeling for image data, though it appears incremental as it builds upon existing continuous normalizing flow methods.
The authors tackled the problem of scaling continuous normalizing flows to higher-resolution images by introducing a multi-resolution variant (MRCNF) that maintains exact likelihood calculation, achieving comparable likelihood values on image datasets with improved performance at higher resolutions and fewer parameters using only 1 GPU.
Recent work has shown that Neural Ordinary Differential Equations (ODEs) can serve as generative models of images using the perspective of Continuous Normalizing Flows (CNFs). Such models offer exact likelihood calculation, and invertible generation/density estimation. In this work we introduce a Multi-Resolution variant of such models (MRCNF), by characterizing the conditional distribution over the additional information required to generate a fine image that is consistent with the coarse image. We introduce a transformation between resolutions that allows for no change in the log likelihood. We show that this approach yields comparable likelihood values for various image datasets, with improved performance at higher resolutions, with fewer parameters, using only 1 GPU. Further, we examine the out-of-distribution properties of (Multi-Resolution) Continuous Normalizing Flows, and find that they are similar to those of other likelihood-based generative models.