HumanOLAT: A Large-Scale Dataset for Full-Body Human Relighting and Novel-View Synthesis

2025-08-12

AI Summary

Existing digital human research is hindered by the lack of high-quality, full-body datasets, particularly those supporting joint modeling of illumination reconstruction and novel-view synthesis. To address this, we introduce the first publicly available large-scale, multi-view, one-light-at-a-time (OLAT) full-body portrait dataset. It employs synchronized multi-view high-dynamic-range (HDR) capture and a systematic OLAT lighting protocol, encompassing ambient illumination, chromatic gradient lighting, and fine-grained OLAT configurations—enabling joint HDR illumination and viewpoint modeling. This dataset fills a critical gap in full-body relighting and novel-view synthesis, significantly enhancing the rigor and comparability of method evaluation. It exposes fundamental challenges in modeling complex human surface–light interactions and establishes a standardized benchmark for advancing algorithmic innovation in neural rendering and photorealistic human synthesis.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Humans and AI: Interaction Techniques and Devices
• Application Domains: Humanities & Computational Social Science

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Security and Privacy: Data transparency and provenance

Abstract

Simultaneous relighting and novel-view rendering of digital human representations is an important yet challenging task with numerous applications. Progress in this area has been significantly limited due to the lack of publicly available, high-quality datasets, especially for full-body human captures. To address this critical gap, we introduce the HumanOLAT dataset, the first publicly accessible large-scale dataset of multi-view One-Light-at-a-Time (OLAT) captures of full-body humans. The dataset includes HDR RGB frames under various illuminations, such as white light, environment maps, color gradients and fine-grained OLAT illuminations. Our evaluations of state-of-the-art relighting and novel-view synthesis methods underscore both the dataset's value and the significant challenges still present in modeling complex human-centric appearance and lighting interactions. We believe HumanOLAT will significantly facilitate future research, enabling rigorous benchmarking and advancements in both general and human-specific relighting and rendering techniques.

Problem

• Lack of public datasets for full-body human relighting and novel-view synthesis
• Challenges in modeling human appearance under complex lighting interactions
• Need for rigorous benchmarking in human-specific relighting techniques

Innovation

• Introduces large-scale HumanOLAT dataset
• Includes multi-view OLAT captures
• Provides HDR RGB frames