4D-DRESS: A 4D Dataset of Real-world Human Clothing with Semantic Annotations

TL;DR

This paper introduces 4D-DRESS, a comprehensive real-world 4D human clothing dataset with detailed annotations, enabling more realistic clothing research and addressing limitations of synthetic datasets.

Contribution

The creation of the first real-world 4D clothing dataset with high-quality scans and a semi-automatic parsing pipeline for accurate annotations.

Findings

Provides 78k textured scans of 64 outfits in diverse motions

Establishes benchmarks for clothing simulation and reconstruction

Demonstrates the dataset's utility in advancing realistic clothing research

Abstract

The studies of human clothing for digital avatars have predominantly relied on synthetic datasets. While easy to collect, synthetic data often fall short in realism and fail to capture authentic clothing dynamics. Addressing this gap, we introduce 4D-DRESS, the first real-world 4D dataset advancing human clothing research with its high-quality 4D textured scans and garment meshes. 4D-DRESS captures 64 outfits in 520 human motion sequences, amounting to 78k textured scans. Creating a real-world clothing dataset is challenging, particularly in annotating and segmenting the extensive and complex 4D human scans. To address this, we develop a semi-automatic 4D human parsing pipeline. We efficiently combine a human-in-the-loop process with automation to accurately label 4D scans in diverse garments and body movements. Leveraging precise annotations and high-quality garment meshes, we establish several benchmarks for clothing simulation and reconstruction. 4D-DRESS offers realistic and challenging data that complements synthetic sources, paving the way for advancements in research of lifelike human clothing. Website: https://ait.ethz.ch/4d-dress.