MeshAvatar: Learning High-quality Triangular Human Avatars from Multi-view Videos

TL;DR

This paper introduces a pipeline for creating high-quality triangular human avatars from multi-view videos, enabling realistic geometry, material decomposition, and support for editing and relighting, overcoming limitations of NeRF-based methods.

Contribution

The method represents avatars with explicit triangular meshes from implicit SDFs and incorporates physics-based rendering and deep supervision for enhanced quality and editability.

Findings

High-quality geometry reconstruction achieved.

Plausible material decomposition demonstrated.

Supports editing, manipulation, and relighting operations.

Abstract

We present a novel pipeline for learning high-quality triangular human avatars from multi-view videos. Recent methods for avatar learning are typically based on neural radiance fields (NeRF), which is not compatible with traditional graphics pipeline and poses great challenges for operations like editing or synthesizing under different environments. To overcome these limitations, our method represents the avatar with an explicit triangular mesh extracted from an implicit SDF field, complemented by an implicit material field conditioned on given poses. Leveraging this triangular avatar representation, we incorporate physics-based rendering to accurately decompose geometry and texture. To enhance both the geometric and appearance details, we further employ a 2D UNet as the network backbone and introduce pseudo normal ground-truth as additional supervision. Experiments show that our method can learn triangular avatars with high-quality geometry reconstruction and plausible material decomposition, inherently supporting editing, manipulation or relighting operations.