TUT Urban Acoustic Scenes 2018

TUT Urban Acoustic Scenes 2018数据集由芬兰坦佩雷理工大学信号处理实验室创建，包含10种不同的城市声场景，如机场、购物中心等，共计24小时的高质量音频。数据集在六个欧洲大城市录制，使用多种设备以增加声学变异性。创建过程涉及在不同城市和场景中选择合适位置进行录音，每位置录音5-6分钟，分割成10秒片段。该数据集主要用于解决声场景分类问题，支持机器学习技术的发展，特别是在处理设备不匹配和声学条件变化方面的研究。

The TUT Urban Acoustic Scenes 2018 dataset was created by the Signal Processing Laboratory of Tampere University of Technology, Finland. It contains 24 hours of high-quality audio covering 10 distinct urban acoustic scenes such as airports, shopping malls, and others. The dataset was recorded in six major European cities using multiple devices to increase acoustic variability. The recording process involved selecting appropriate locations across different cities and scenes, with each location being recorded for 5 to 6 minutes, and the audio was then segmented into 10-second clips. This dataset is primarily utilized for addressing acoustic scene classification tasks, supporting the advancement of machine learning technologies, particularly research focusing on handling device mismatch and variations in acoustic conditions.