TAU-NIGENS Spatial Sound Events 2021

TAU-NIGENS Spatial Sound Events 2021数据集由坦佩雷大学音频研究组创建，包含800个一分钟的立体声录音，用于声音事件定位和检测。数据集基于真实房间的空间房间脉冲响应（SRIRs），模拟了静态或移动声音事件在混响和环境噪声条件下的录音。数据集内容包括12类目标声音事件和3类方向干扰器，共约900个声音样本。创建过程中，使用了13个房间的SRIRs，通过随机抽取和插值合成移动声音事件。该数据集主要用于机器人听觉、机器听觉监测、智能家居环境等领域的研究，旨在解决复杂环境下的声音事件定位和检测问题。

The TAU-NIGENS Spatial Sound Events 2021 dataset was developed by the Audio Research Group of Tampere University. It comprises 800 one-minute stereo recordings designed for sound event localization and detection tasks. Built upon spatial room impulse responses (SRIRs) acquired from real-world rooms, this dataset simulates recordings of static or moving sound events under reverberant and ambient noise conditions. The dataset includes 12 categories of target sound events and 3 types of directional interferers, with a total of approximately 900 sound samples. During its construction, SRIRs from 13 rooms were employed, and moving sound events were synthesized through random sampling and interpolation. This dataset is primarily applied to research in fields such as robotic audition, machine auditory monitoring, and smart home environments, with the goal of solving sound event localization and detection problems in complex real-world scenarios.