HUMBI: A Large Multiview Dataset of Human Body Expressions and Benchmark Challenge

TL;DR

HUMBI is a comprehensive multiview dataset capturing human body expressions with natural clothing, enabling advanced 3D modeling and a benchmark challenge for pose-guided appearance rendering to enhance social telepresence.

Contribution

The paper introduces HUMBI, a large-scale multiview dataset with synchronized HD cameras, and establishes a new benchmark for 3D human expression modeling and rendering.

Findings

HUMBI effectively supports learning and reconstructing detailed human models.

It complements existing datasets with diverse subjects and views.

The benchmark advances photorealistic 3D human expression rendering.

Abstract

This paper presents a new large multiview dataset called HUMBI for human body expressions with natural clothing. The goal of HUMBI is to facilitate modeling view-specific appearance and geometry of five primary body signals including gaze, face, hand, body, and garment from assorted people. 107 synchronized HD cameras are used to capture 772 distinctive subjects across gender, ethnicity, age, and style. With the multiview image streams, we reconstruct high fidelity body expressions using 3D mesh models, which allows representing view-specific appearance. We demonstrate that HUMBI is highly effective in learning and reconstructing a complete human model and is complementary to the existing datasets of human body expressions with limited views and subjects such as MPII-Gaze, Multi-PIE, Human3.6M, and Panoptic Studio datasets. Based on HUMBI, we formulate a new benchmark challenge of a pose-guided appearance rendering task that aims to substantially extend photorealism in modeling diverse human expressions in 3D, which is the key enabling factor of authentic social tele-presence. HUMBI is publicly available at http://humbi-data.net