Abstract
This paper describes Task 2 of the DCASE 2018 Challenge, titled "General-purpose audio tagging of Freesound content with AudioSet labels". This task was hosted on the Kaggle platform as "Freesound General-Purpose Audio Tagging Challenge". The goal of the task is to build an audio tagging system that can recognize the category of an audio clip from a subset of 41 heterogeneous categories drawn from the AudioSet Ontology. We present the task, the dataset prepared for the competition, and a baseline system.
Abstract (translated)
本文描述了DCASE 2018 Challenge的任务2,标题为“具有AudioSet标签的Freesound内容的通用音频标记”。此任务作为“Freesound通用音频标记挑战”在Kaggle平台上托管。该任务的目标是构建一个音频标记系统,该系统可以从AudioSet Ontology中提取的41个异类分类的子集中识别音频剪辑的类别。我们提出了任务,为比赛准备的数据集和基线系统。
URL
https://arxiv.org/abs/1807.09902