TUT Urban Acoustic Scenes Mobile 2018

本文介绍了DCASE 2018挑战的声学场景分类任务和为任务提供的TUT城市声学场景2018数据集，并评估了基线系统在任务中的性能。与挑战的前几年一样，该任务被定义为使用监督的封闭集分类设置将短音频样本分类为预定义的声学场景类别之一。新记录的TUT城市声学场景2018数据集由十个不同的声学场景组成，并记录在六个欧洲大城市中，因此它具有比以前用于此任务的数据集更高的声学可变性，并且除了高质量的双耳记录之外，它还包括用移动设备记录的数据。我们还介绍了由卷积神经网络组成的基线系统及其使用推荐的交叉验证设置在子任务中的性能。

This paper introduces the acoustic scene classification task of the DCASE 2018 Challenge and the TUT Urban Acoustic Scenes 2018 dataset provided for the task, and evaluates the performance of the baseline system for this task. Similar to previous editions of the challenge, this task is defined as a supervised closed-set classification setup to categorize short audio samples into one of the predefined acoustic scene categories. The newly recorded TUT Urban Acoustic Scenes 2018 dataset consists of ten distinct acoustic scenes, recorded in six major European cities, thus it features higher acoustic variability than datasets previously used for this task, and in addition to high-quality binaural recordings, it also includes data captured with mobile devices. We also introduce the baseline system composed of convolutional neural networks and its performance in the subtasks using the recommended cross-validation setup.

搜集汇总

背景与挑战

背景概述

TUT Urban Acoustic Scenes Mobile 2018 数据集是为DCASE 2018挑战赛的声学场景分类任务而构建。它收录了在六个欧洲大城市记录的十个不同声学场景的音频，包含移动设备及双耳录音，旨在提供比以往数据集更高的声学可变性。