AudioCaps

AudioCaps 是一个带有事件描述的声音数据集，它是为音频字幕任务引入的，声音来自 AudioSet 数据集。为注释者提供了音轨和类别提示（如果需要，还提供额外的视频提示）。

AudioCaps is an audio dataset equipped with event descriptions, which was introduced for the audio captioning task. Its audio materials are sourced from the AudioSet dataset. Annotators were provided with audio tracks and category prompts, with additional video prompts available if necessary.

搜集汇总

背景与挑战

背景概述

AudioCaps是一个用于音频字幕任务的声音数据集，包含来自AudioSet数据集的事件描述。该数据集由首尔国立大学于2019年发布，为注释者提供了音轨和类别提示以辅助标注。