Hybrid Generative Models for Two-Dimensional Datasets
This addresses a gap in generative modeling for scientific and other domains with continuous-valued 2D data, offering a generalizable method, though it appears incremental as it builds on existing generative models and representation bases.
The paper tackles the problem of generative modeling for two-dimensional datasets beyond conventional images, where existing methods fail to capture correlations and handle continuous values, by proposing a novel approach that moves computations to representation bases, demonstrating its effectiveness on imaging and scientific computing datasets with comprehensive performance comparisons and a new evaluation metric.
Two-dimensional array-based datasets are pervasive in a variety of domains. Current approaches for generative modeling have typically been limited to conventional image datasets and performed in the pixel domain which do not explicitly capture the correlation between pixels. Additionally, these approaches do not extend to scientific and other applications where each element value is continuous and is not limited to a fixed range. In this paper, we propose a novel approach for generating two-dimensional datasets by moving the computations to the space of representation bases and show its usefulness for two different datasets, one from imaging and another from scientific computing. The proposed approach is general and can be applied to any dataset, representation basis, or generative model. We provide a comprehensive performance comparison of various combinations of generative models and representation basis spaces. We also propose a new evaluation metric which captures the deficiency of generating images in pixel space.