LayoutFlow: Flow Matching for Layout Generation
This addresses layout generation for graphic design applications, offering a faster alternative to diffusion-based models, but it is incremental as it adapts flow matching to an existing task.
The paper tackled layout generation by proposing LayoutFlow, a flow-based model that moves elements from an initial sample to a final prediction, achieving performance on par with state-of-the-art models while being significantly faster.
Finding a suitable layout represents a crucial task for diverse applications in graphic design. Motivated by simpler and smoother sampling trajectories, we explore the use of Flow Matching as an alternative to current diffusion-based layout generation models. Specifically, we propose LayoutFlow, an efficient flow-based model capable of generating high-quality layouts. Instead of progressively denoising the elements of a noisy layout, our method learns to gradually move, or flow, the elements of an initial sample until it reaches its final prediction. In addition, we employ a conditioning scheme that allows us to handle various generation tasks with varying degrees of conditioning with a single model. Empirically, LayoutFlow performs on par with state-of-the-art models while being significantly faster.