Efficient Multi-view Performance Capture of Fine-Scale Surface Detail

TL;DR

This paper introduces a novel method for capturing fine-scale surface details in deforming meshes from multi-view video, improving detail accuracy over existing coarse-to-medium scale models.

Contribution

It proposes an implicit surface representation using 3D and 2D Gaussians, enabling efficient, stable, and detailed surface deformation optimization without correspondence or sampling issues.

Findings

Robustly captures more fine-scale detail than related methods.

Handles occlusions implicitly with smooth energy formulation.

Qualitative and quantitative validation on human subjects.

Abstract

We present a new effective way for performance capture of deforming meshes with fine-scale time-varying surface detail from multi-view video. Our method builds up on coarse 4D surface reconstructions, as obtained with commonly used template-based methods. As they only capture models of coarse-to-medium scale detail, fine scale deformation detail is often done in a second pass by using stereo constraints, features, or shading-based refinement. In this paper, we propose a new effective and stable solution to this second step. Our framework creates an implicit representation of the deformable mesh using a dense collection of 3D Gaussian functions on the surface, and a set of 2D Gaussians for the images. The fine scale deformation of all mesh vertices that maximizes photo-consistency can be efficiently found by densely optimizing a new model-to-image consistency energy on all vertex positions. A principal advantage is that our problem formulation yields a smooth closed form energy with implicit occlusion handling and analytic derivatives. Error-prone correspondence finding, or discrete sampling of surface displacement values are also not needed. We show several reconstructions of human subjects wearing loose clothing, and we qualitatively and quantitatively show that we robustly capture more detail than related methods.