HEADSET: Human Emotion Awareness under Partial Occlusions Multimodal DataSET

HEADSET数据集是由Technische Universität Berlin、INRIA、Tampere University和Ernst-Abbe University of Applied Sciences Jena共同创建的多模态数据库，专注于在部分遮挡下的人类情感意识，特别是佩戴头戴式显示器（HMD）的情况。该数据集包含27名参与者的数据，通过配备同步RGB和深度摄像机的体积捕捉工作室以及Lytro Illum相机捕获的光场数据进行采集。数据集内容丰富，包括彩色3D点云、纹理3D网格、多视角RGB-D图像和光场图像，旨在为沉浸式媒体技术和XR应用中的真实人类交互提供高质量测试集。

The HEADSET dataset is a multimodal database co-created by Technische Universität Berlin, INRIA, Tampere University, and Ernst-Abbe University of Applied Sciences Jena. It focuses on human emotional awareness under partial occlusion, particularly in scenarios where users wear Head-Mounted Displays (HMDs). The dataset contains data from 27 participants, collected via a volumetric capture studio equipped with synchronized RGB and depth cameras, as well as light field data captured by a Lytro Illum camera. It includes rich content such as colored 3D point clouds, textured 3D meshes, multi-view RGB-D images and light field images, aiming to provide high-quality test sets for real-world human interaction in immersive media technology and XR applications.

搜集汇总

背景与挑战

背景概述

HEADSET是一个多模态数据库，专注于在部分遮挡（如佩戴头戴式显示器）下的人类情感意识，由多所大学联合创建。它包含27名参与者的彩色3D点云、纹理3D网格、多视角RGB-D图像和光场图像，旨在为沉浸式媒体技术和XR应用提供高质量的真实人类交互测试集。