Enhanced 3D Shape Analysis via Information Geomerty

Amit Vishwakarma, K.S. Subrahamanian Moosath · 2025

Read Enhanced 3D Shape Analysis via Information Geomerty by Amit Vishwakarma, K.S. Subrahamanian Moosath (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-16213.txt.

Introduction

Amit Vishwakarma and K S Subrahamanian Moosath compare 3D shapes in a new way in an arXiv paper posted in 2025. Point clouds, clouds of measured points on a surface, drive graphics, vision, and robotics. Standard distance measures between clouds miss global structure and jump wildly when outliers appear.

The authors model each cloud as a mixture of Gaussians living on a curved statistical space, using information geometry. They prove that space is a proper statistical manifold and define a modified symmetric divergence between mixtures. Unlike older approximations, their measure is provably bounded above and below, so it never blows up numerically. On human pose data and animal shape data, the measure varied smoothly with real geometric change and beat classic distances and earlier divergence approximations.

Catalog record

CREATORAmit Vishwakarma, K.S. Subrahamanian Moosath
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-16213

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Enhanced 3D Shape Analysis via Information Geomerty Amit Vishwakarma, K.S. Subrahamanian Moosath Abstract Three-dimensional point clouds provide highly accurate digital representations of objects, essential for applications in computer graphics, photogrammetry, computer vision, and robotics. However, comparing point clouds faces significant challenges due to their unstructured nature and the complex geometry of the surfaces they represent. Traditional geometric metrics such as Hausdorff and Chamfer distances often fail to capture global statistical structure and exhibit sensitivity to outliers, while existing Kullback-Leibler (KL) divergence approximations for Gaussian Mixture Models can produce unbounded or numerically unstable values. This paper introduces an information geometric framework for 3D point cloud shape analysis by representing point clouds as Gaussian Mixture Models (GMMs) on a statistical manifold. We prove that the space of GMMs forms a statistical manifold and propose the Modified Symmetric Kullback-Leibler (MSKL) divergence with theoretically guaranteed upper and lower bounds, ensuring numerical stability for all GMM comparisons. Through comprehensive experiments on human pose discrimination (MPI-FAUST dataset) and animal shape comparison (G-PCD dataset), we demonstrate that MSKL provides stable and monotonically varying values that directly reflect geometric variation, outperforming traditional distances and existing KL approximations. keywords Information Geometry, Point Cloud, Gaussian Mixture Model, Statistical Manifold, Divergence 1 Introduction Three-dimensional point clouds provide highly accurate digital representations of objects which makes it important for the applications in computer graphics, photogrammetry, computer vision, construction, and remote sensing 10.1016/j.aei.2019.02.007 ; chen2022stpls3d ; akagic2022cv . Recent work has addressed various geometric analysis tasks including robust symmetry detection in the presence of outliers Nagar2025 . Extracting meaningful information from point clouds is fundamental for shape analysis tasks such as pose estimation, deformation tracking, registration, retrieval and autoencoding kucak2022ransac ; zhang2025dlpc ; huang2021pcsurvey . However, comparing point clouds faces unique challenges due to their unstructured nature and the complex geometry of the surfaces they represent. Unlike structured grid data with natural Euclidean metrics, point clouds lack a common coordinate system, making direct comparisons difficult. When comparing shapes sampled from 3D surfaces, the objective is to capture global geometric similarity rather than point-wise correspondence. Classical geometric distances such as Hausdorff hausdorff2008gesammelte and Chamfer yang2018 measure spatial proximity between point sets. While effective for certain tasks, they often fail to capture broader statistical structure, exhibit sensitivity to outliers and sampling irregularities, and do not characterize the global distribution of points. Earth Mover’s Distance (EMD) rubner2000earth computes similarity as the minimum cost of transforming one distribution into another, providing a transport-based metric. However, EMD requires solving an expensive optimization problem that scales poorly with point cloud size and often struggles to capture intrinsic geometric relationship between the shapes. Recent works have explored probabilistic representations using GMMs for point cloud registration and comparison jian2005robust . While these methods make use of statistical modeling, they generally rely on approximations of the Kullback–Leibler (KL) divergence and these approximations have some limitations: • KL Weighted-Average (KL {}_{\text{WA}} ) 1238387 ; hershey2007approximating approximation provides a lower bound by averaging pairwise KL divergences between Gaussian components weighted by mixing coefficients. However, this approximation can significantly underestimate the true divergence, particularly when mixture components have poor overlap. • KL Matching-Based (KL {}_{\text{MB}} ) 1238387 ; hershey2007approximating approximation matches each component of one mixture to the closest component in another, yielding an upper bound. This matching strategy can overestimate divergence and is sensitive to the number and initialization of components. • Furthermore, the standard KL divergence is asymmetric ( {\text{KL}}(p||q)\neq{\text{KL}}(q||p) ) and can be unbounded when comparing GMMs, leading to numerical instabilities and undefined comparisons in practical scenarios. This work develops an information geometric framework for stable shape similarity measurement by representing each point cloud as a GMM defined on a low-dimensional latent space. The framework includes preprocessing, extraction of local descriptors, manifold embedding via Isomap, and GMM fitting through the Expectation–Maximization (EM) algorithm. With this representation, the collection of all such GMMs forms a statistical manifold, enabling the use of divergence-based tools for shape analysis. We introduce the Modified Symmetric KL (MSKL) divergence which addresses the instability issue of exisitng KL approaximations. The theoretical properties, including bounds and stability for MSKL are established in the subsequent sections. In addition, we consider two well-known approximations for KL-divergence on GMMs—the weighted-average ( KL_{\text{WA}} ) and matching-based (KL {}_{\text{MB}} ) which provide complementary baselines frequently used in mixture model comparison. The main contributions of this work are: • Established a rigorous mathematical framework demonstrating that point clouds represented as GMMs form a statistical manifold, enabling the use of information-geometric methods. • Introduced the MSKL divergence and prove that it is well defined by establishing upper and lower bounds, providing theoretical guarantee for GMM-based shape comparison. • Compared the performance of MSKL with symmetric KL-based approximations (SKL {}_{\text{WA}} , SKL {}_{\text{MB}} ) and classical geometric distances (Hausdorff, Chamfer). To demonstrate the practical performance of the framework, we compare MSKL with Symmetric KL divergence (SKL), SKL {}_{\text{WA}} , SKL {}_{\text{MB}} and the traditional distances Hausdorff, and Chamfer across two shape categories: (i) human shapes from the MPI FAUST dataset mpi_faust_dataset and (ii) animal shapes from the G-PCD dataset epfl_geometry_point_cloud_dataset . These experiments highlight how statistical manifold representations capture both local geometry and global structure more effectively than metrics. 2 Related Works In this section, we review existing methods for comparing point clouds, which can be broadly classified into two categories: metric-based distances and divergence-based statistical approaches. 2.1 Metric-Based Approaches 1. Hausdorff Distance. For two finite points P,Q\in\mathbb{R}^{m} the Hausdorff distance hausdorff2008gesammelte is defined as D_{H}(P,Q)=\max\left\{\sup_{p\in P}\inf_{q\in Q}\|p-q\|,\sup_{q\in Q}\inf_{p\in P}\|p-q\|\right\}. (1) This measure emphasizes the largest difference between two sets of 3D points, which makes it particularly sensitive to unusual points that do not fit in the pattern. 2. Chamfer Distance. The Chamfer distance yang2018 provides an alternative that averages point-to-point distances, thereby reducing the impact of outliers. It is widely used in 3D reconstruction and mesh generation. Ch(P,Q)=\frac{1}{|P|}\sum_{p\in P}\min_{q\in Q}\|p-q\|+\frac{1}{|Q|}\sum_{q\in Q}\min_{p\in P}\|p-q\|. (2) This method calculates the average of the squared distances between each point in one set to its nearest point in the other set, offering a balance between sensitivity to large discrepancies and outliers. In this method differences in point cloud shape and distribution is not taken care of. 3. Earth Mover’s Distance. The Earth Mover’s Distance (EMD) formulates point cloud comparison as an optimal transport problem: \text{EMD}(A,B)=\min_{\phi:A\leftrightarrow B}\sum_{a\in A}\|a-\phi(a)\|, (3) where \phi denotes a bijection between point clouds A and B . While EMD provides a principled transport-based metric, it requires solving an expensive optimization problem that scales poorly with point cloud size ( O(n^{3}\log n) ), making it impractical for large-scale applications rubner2000earth . 4. Gromov-Hausdorff Distance. In hausdorff2008gesammelte and yang2018 comparing two objects reduced to the comparison between the corresponding interpoint distance matrices in the multidimensional scaling. These methods primarily capture rigid similarities and do not account for isometric objects. To address this limitation Facundo Mémoli and Guillermo Sapiro memoli2004comparing introduced a method based on the Gromov-Hausdorff distance, which provides a theoretical framework for isometric-invariant recognition. The Gromov-Hausdorff distance, d_{GH} measures the similarity between two metric spaces X and Y by embedding them into a common metric space Z and minimizing the Hausdorff distance d_{H}^{Z}(X,Y) . The Gromov-Hausdorff distance is defined as d_{GH}(X,Y)=\inf_{Z,f,g}d_{H}^{Z}(f(X),g(Y)), (4) where, d_{H}^{Z}(f(X),g(Y))=\max\left(\sup_{x\in f(X)}d_{Z}(x,g(Y)),\sup_{y\in g(Y)}d_{Z}(y,f(X))\right). Here, f and g

Read the complete text