DCASE 2022 Challenge

此子任务涉及使用音频信号的声音内容文本描述 (即音频标题) 来检索音频信号。人类书面音频标题将用作文本查询。对于每个文本查询，此任务的目标是从给定的数据集检索10个音频文件，并根据它们与查询的匹配对其进行排序。通过此子任务，我们旨在激发对具有不受约束的文本描述的基于语言的音频检索的进一步研究。This subtask involves retrieving audio signals using textual descriptions of their acoustic content, i.e., audio captions. Human-written audio captions will serve as textual queries. For each textual query, the goal of this task is to retrieve 10 audio files from the given dataset and rank them based on their relevance to the query. Through this subtask, we aim to promote further research into language-based audio retrieval with unconstrained textual descriptions.

搜集汇总

背景与挑战

背景概述

DCASE 2022 Challenge数据集专注于基于语言的音频检索任务，通过人类书写的音频标题作为文本查询，从数据集中检索并排序匹配的音频文件。该数据集由坦佩雷大学于2022年发布，旨在推动不受约束文本描述的音频检索研究。