Naturalistic audio-visual volumetric sequences dataset of sounding actions

本数据集名为‘Naturalistic audio-visual volumetric sequences dataset of sounding actions’，由德国弗劳恩霍夫IIS和英国萨里大学CVSSP联合创建。数据集包含40个短动作序列，涵盖多种声音类型和特征，旨在为沉浸式和交互式研究提供高质量的音频和视频数据。数据集通过精心设计的录音技术捕捉，包括麦克风技术、位置和水平，以及背景噪声和混响效果的考虑。该数据集适用于研究跨模态感知影响，特别是在虚拟环境中的应用，如VR/AR/XR中的6DoF交互，旨在解决音频和视频数据在技术研究和科学感知研究中的集成测试问题。

This dataset, entitled 'Naturalistic audio-visual volumetric sequences dataset of sounding actions', was co-created by Fraunhofer IIS of Germany and CVSSP, University of Surrey, UK. It comprises 40 short action sequences spanning a wide range of sound types and acoustic characteristics, with the goal of supplying high-quality audio and video data for immersive and interactive research. The dataset was acquired via meticulously designed recording workflows, which incorporate microphone technologies, placement, and recording levels, while also accounting for background noise and reverberation effects. This dataset supports research on cross-modal perceptual impacts, particularly for applications in virtual environments such as 6DoF interactions within VR/AR/XR, and aims to address the integration testing challenges of audio and video data in both technical research and scientific perceptual studies.