VoxCeleb1

该数据集包含了来自1251位发言者的语音数据，这些数据在背景噪声的环境下录制。数据集被划分为开发集和评估集两部分。规模上，它包含了1211位发言者的语音，总时长达到340小时，旨在用于说话人识别任务。

This dataset contains speech data from 1251 speakers, which was recorded under background noise conditions. The dataset is split into two subsets: a development set and an evaluation set. In terms of scale, it encompasses speech from 1211 speakers with a total duration of 340 hours, and it is designed for speaker recognition tasks.

搜集汇总

背景与挑战

背景概述

VoxCeleb1是一个包含1,251位名人超过10万条语音片段的大规模说话人识别数据集，数据来源于YouTube视频，适用于说话人验证和识别任务。数据集提供了详细的验证和识别分割，但原始音频和元数据已不再公开提供。