VisDiff: SDF-Guided Polygon Generation for Visibility Reconstruction and Recognition

TL;DR

VisDiff is a diffusion-based method that reconstructs polygons from visibility graphs by estimating the signed distance function, enabling better learning of visibility relationships and outperforming existing methods.

Contribution

We introduce VisDiff, a novel diffusion approach that uses SDF estimation for polygon reconstruction from visibility graphs, addressing limitations of previous methods.

Findings

Achieved 26% improvement in F1-Score over standard methods.

Demonstrated effective learning of visibility relationships.

Provided a curated dataset for benchmarking.

Abstract

The ability to capture rich representations of combinatorial structures has enabled the application of machine learning to tasks such as analysis and generation of floorplans, terrains, images, and animations. Recent work has primarily focused on understanding structures with well-defined features, neighborhoods, or underlying distance metrics, while those lacking such characteristics remain largely unstudied. Examples of these combinatorial structures can be found in polygons, where a small change in the vertex locations causes a significant rearrangement of the combinatorial structure, expressed as a visibility or triangulation graphs. Current representation learning approaches fail to capture structures without well-defined features and distance metrics. In this paper, we study the open problem of Visibility Reconstruction: Given a visibility graph $G$, construct a polygon $P$ whose visibility graph is $G$. We introduce VisDiff, a novel diffusion-based approach to generate polygon $P$ from the input visibility graph $G$. The main novelty of our approach is that, rather than generating the polygon's vertex set directly, we first estimate the signed distance function (SDF) associated with the polygon. The SDF is then used to extract the vertex location representing the final polygon. We show that going through the SDF allows VisDiff to learn the visibility relationship much more effectively than generating vertex locations directly. In order to train VisDiff, we create a carefully curated dataset. We use this dataset to benchmark our method and achieve 26% improvement in F1-Score over standard methods as well as state of the art approaches.