Learned Vertex Descent: A New Direction for 3D Human Model Fitting

TL;DR

Learned Vertex Descent (LVD) introduces a neural network-based optimization method for 3D human model fitting that predicts vertex descent directions, enabling fast and accurate fitting of clothed human bodies from images and scans.

Contribution

We develop a neural network ensemble that predicts vertex descent directions, enabling rapid and accurate 3D human model fitting through gradient descent optimization.

Findings

Achieves significant improvement over state-of-the-art in 3D human body fitting.

Fitting process converges in a fraction of a second.

Effective for both human and hand model fitting.

Abstract

We propose a novel optimization-based paradigm for 3D human model fitting on images and scans. In contrast to existing approaches that directly regress the parameters of a low-dimensional statistical body model (e.g. SMPL) from input images, we train an ensemble of per-vertex neural fields network. The network predicts, in a distributed manner, the vertex descent direction towards the ground truth, based on neural features extracted at the current vertex projection. At inference, we employ this network, dubbed LVD, within a gradient-descent optimization pipeline until its convergence, which typically occurs in a fraction of a second even when initializing all vertices into a single point. An exhaustive evaluation demonstrates that our approach is able to capture the underlying body of clothed people with very different body shapes, achieving a significant improvement compared to state-of-the-art. LVD is also applicable to 3D model fitting of humans and hands, for which we show a significant improvement to the SOTA with a much simpler and faster method.