Abstract:Intrinsic symmetry detection, phrased as finding intrinsic self-isometries, courts much attention in recent years. However, extracting dense global symmetry from the shape undergoing moderate non-isometric deformations is still a challenge to the state-of-the-art methods. To tackle this problem, we develop an automatic and robust global intrinsic symmetry detector based on functional maps. The main challenges of applying functional maps lie in how to amend the previous numerical solution scheme and construct reliable and enough constraints. We address the first challenge by formulating the symmetry detection problem as an objective function with descriptor, regional and orthogonality constraints and solving it directly. Compared with refining the functional map by a post-processing, our approach does not break existing constraints and generates more confident results without sacrificing efficiency. To conquer the second challenge, we extract a sparse and stable symmetry-invariant point set from shape extremities and establish symmetry electors based on the transformation, which is constrained by the symmetric point pairs from the set. These electors further cast votes on candidate point pairs to extract more symmetric point pairs. The final functional map is generated with regional constraints constructed from the above point pairs. Experimental results on TOSCA and SCAPE Benchmarks show that our method is superior to the state-of-the-art methods. Graphical abstractDisplay Omitted HighlightsA robust global intrinsic symmetry method based on functional maps is proposed.We employ an optimization with orthogonality constraints and solve it directly.An reliable symmetry-invariant set is extracted to establish initial constraints.A voting scheme is proposed to establish more regional constraints.A considerable amount of comparisons with the state-of-the-art are provided.

Symmetry and Orbit Detection Via Lie-Algebra Voting.

Multiresolution Rotational Symmetry Detection Via Radius-Based Frieze-Expansion.

Coupled Structure-from-motion and 3D Symmetry Detection for Urban Facades

LieDetect: Detection of representation orbits of compact Lie groups from point clouds

Robust Symmetry Detection via Riemannian Langevin Dynamics

SymmetryLens: A new candidate paradigm for unsupervised symmetry learning via locality and equivariance

Robust discovery of partial rigid symmetries on 3D models.

Deep Learning Symmetries and Their Lie Groups, Algebras, and Subalgebras from First Principles

Symmetry Discovery Beyond Affine Transformations

3D Symmetry Detection by a Single Image and Geometric Transformation

Polygon Recognition and Symmetry Detection

A Hierarchical Symmetry Detection Algorithm Based on Voxelization

Mirror Symmetry for Special Nilpotent Orbit Closures

A Unified Framework to Enforce, Discover, and Promote Symmetry in Machine Learning

Properly Constrained Orthonormal Functional Maps for Intrinsic Symmetries.

Symmetry Discovery for Different Data Types

On Symmetry and Multiple-View Geometry: Structure, Pose, and Calibration from a Single Image

Symmetry-aware Graph Matching.

Using Machine Learning to Detect Rotational Symmetries from Reflectional Symmetries in 2D Images

Classification of Affine Symmetry Groups of Orbit Polytopes

Automatic and Exact Symmetry Recognition of Structures Exhibiting High-Order Symmetries