TELA: Text to Layer-wise 3D Clothed Human Generation

TL;DR

This paper introduces TELA, a novel layer-wise 3D human generation method from text that enables fine-grained clothing editing and disentangled modeling through progressive optimization and stratified rendering.

Contribution

The paper proposes a layer-wise representation and progressive optimization strategy for 3D clothed human generation, improving control and disentanglement over previous holistic models.

Findings

Achieves state-of-the-art 3D clothed human generation quality.

Supports clothing editing applications like virtual try-on.

Provides high-quality clothing disentanglement for detailed garment control.

Abstract

This paper addresses the task of 3D clothed human generation from textural descriptions. Previous works usually encode the human body and clothes as a holistic model and generate the whole model in a single-stage optimization, which makes them struggle for clothing editing and meanwhile lose fine-grained control over the whole generation process. To solve this, we propose a layer-wise clothed human representation combined with a progressive optimization strategy, which produces clothing-disentangled 3D human models while providing control capacity for the generation process. The basic idea is progressively generating a minimal-clothed human body and layer-wise clothes. During clothing generation, a novel stratified compositional rendering method is proposed to fuse multi-layer human models, and a new loss function is utilized to help decouple the clothing model from the human body. The proposed method achieves high-quality disentanglement, which thereby provides an effective way for 3D garment generation. Extensive experiments demonstrate that our approach achieves state-of-the-art 3D clothed human generation while also supporting cloth editing applications such as virtual try-on. Project page: http://jtdong.com/tela_layer/