Searcharxiv⌕ Search

arXiv · 2610.31892

VisFlow: Interpretable Flow Matching for Visibility Reconstruction, Characterization, and Recognition

Abstract

Methods for learning continuous representations, such as latent manifolds, are fundamental and ubiquitous in modern machine learning. Yet many geometric objects combine continuous shape information with discrete combinatorial structure. This combination presents a fundamental challenge for representation learning because small geometric perturbations can induce abrupt changes in the underlying discrete representation. As a result, the smoothness assumptions that make continuous representations effective no longer hold. Polygons provide a canonical example. Their vertex locations vary continuously, but infinitesimal changes in geometry can dramatically alter the associated visibility graph. Existing methods address the inverse mapping from visibility graphs to polygon geometry using vertex-, triangulation-, or signed-distance-function (SDF)-based representations. Vertex- and triangulation-based approaches construct polygons through explicit geometric operations. However, they are less effective at modeling the relationship between geometry and visibility. SDF-based approaches capture this relationship more naturally. Their intermediate states, however, do not correspond to explicit stages of polygon construction. We present VisFlow, a novel, interpretable flow formulation that combines continuous SDF representations with explicit polygon evolution. We evaluate VisFlow on Visibility Reconstruction and Visibility Characterization. Visibility Reconstruction generates one polygon consistent with a visibility graph (G), while Visibility Characterization characterizes all such polygons. VisFlow maintains 20% more visibility relations on Visibility Reconstruction and improves coverage by 30% on Visibility Characterization compared to state-of-the-art methods.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Rahul Moorthy, Volkan Isler. 2026-10-22. VisFlow: Interpretable Flow Matching for Visibility Reconstruction, Characterization, and Recognition. https://searcharxiv.com/abs/2610.31892

Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.

KEEP EXPLORING

Related papers

Smooth Surfaces via Nets of Geodesics

We study the reconstruction of smooth surfaces from networks of spatial curves prescribed as geodesics. Motivated by applications in computer-aided geometric design and the perception of three-dimensional shape, we propose a variational approach minimizing the integral of squared Gaussian curvature over surfaces spanning geodesic contours. Using isothermal coordinates, we derive the Euler-Lagrange equation for the conformal factor of the metric and the associated Dirichlet and Neumann boundary conditions. We also consider a biharmonic approximation obtained by linearizing the nonlinear equation. Numerical experiments are performed on four quadrilateral cells of a geodesic network constructed on an ellipsoid. We compute approximate isothermal coordinates, solve the biharmonic boundary-value problem, and compare the resulting curvature with that of the reference surface. These experiments consist of an important step towards a complete framework for the computation of smooth surfaces from geodesic curves.

cs.CG↗

A Linear-Size Well-Separated Pair Decomposition for Unit Disk Graphs

Given a metric space $(X, d)$ a $c$-well separated pair decomposition (WSPD) is a collection of `dumbbell' pairs of subsets $(A, B)$, $A, B\subseteq X$, where the distance between the two sets is at least $c$ times the diameter of each set and every pair of points in $X$ is covered by one such pair. It is known that for a weighted unit disk graph metric of $n$ points, there is a WSPD of size $O(n\log n)$. But it is open whether this bound can be improved to $O(n)$. In this note, we answer this question affirmatively and provide an algorithm that yields a linear-size WSPD for unit disk graphs.

cs.CG↗

An Omnidirectional WECT: Closed-Form Computation of the Directionally Integrated Weighted Euler Characteristic Transform

The weighted Euler characteristic transform (WECT) and related directional descriptors have become popular, efficiently computable tools for shape analysis across a variety of domains. In practice, however, they are approximated over a finite set of filtration directions, and thus the vectorized descriptor depends on this choice of directions and the rotational orientation of the data. In this manuscript, we formally introduce the directionally integrated weighted Euler characteristic transform (DIWECT), which averages the WECT over all directions $s \in \mathbb{S}^n$ for a weighted simplicial complex immersed in $\mathbb{R}^{n+1}$. A framework for computing the DIWECT exactly, without sampling directions, is outlined: for each simplex, we partition $\mathbb{S}^n$ into the sets of directions in which a given vertex enters the filtration last, and then the DIWECT at height $t$ reduces to measuring intersections of these regions with a set of directions in which the simplex has entered the filtration by height~$t$. We provide implementations for complexes immersed in $\mathbb{R}^2$ and $\mathbb{R}^3$, in which case the problem reduces to computing arc lengths and spherical areas. Experiments on synthetic and real-world data show that the DIWECT, while resulting in more compact vectorizations, also successfully encodes shape information and discards noise from data orientation. On a barley seed classification task, it outperforms the WECT in every pose except a hand-curated frame, whose advantage disappears under a two-degree rotation error. Compared with averaging the WECT over sampled directions, the DIWECT is often faster at matched accuracy. In one instance, this is upwards of a $50 \times$ speedup on a one-million-cell image with an error of $10^{-3}$.

cs.CG↗