Device Recorded VCTK (DR-VCTK)

DR-VCTK is device recorded version of VCTK dataset on common consumer devices (laptop, tablet and smartphone) in office environment. We name it as Device Recorded VCTK (DR-VCTK). This repository contains the complete DR-VCTK dataset. A small subset of DR-VCTK is available at https://doi.org/10.7488/ds/2316. Clean VCTK audio files and orthographic transcriptions are publicly available here: https://doi.org/10.7488/ds/2645 This dataset contains 109 English speakers with different accents. There are around 400 sentences available from each speaker. For this recording, 8 different microphones were used. Type and quality of the selected microphones are shown in the dataset structure section. Bose 404600 was used as a high quality speaker and it was set in the 2 meters distance from the microphones. Recording was done in the medium size office in two background noise conditions (when the window was opened or closed).

DR-VCTK是VCTK数据集在办公环境下，通过普通消费级设备（笔记本电脑、平板电脑与智能手机）录制的设备录制版本，我们将其命名为设备录制版VCTK（Device Recorded VCTK，缩写为DR-VCTK）。本仓库包含完整的DR-VCTK数据集，该数据集的小型子集可通过https://doi.org/10.7488/ds/2316获取。纯净版VCTK音频文件与正字法转录文本可通过https://doi.org/10.7488/ds/2645公开获取。该数据集涵盖109名带有不同口音的英语说话者，每位说话者约提供400句语音语句。本次录制共使用8款不同的麦克风，所选麦克风的型号与质量详情可参见数据集结构章节。本次录制采用Bose 404600作为高品质扬声器，并将其设置在距离麦克风2米的位置。录制场地为一间中型办公室，共设置两种背景噪声条件（窗户开启或关闭状态）。

搜集汇总

背景与挑战

背景概述

DR-VCTK是一个在办公室环境中使用常见消费设备录制的英语语音数据集，包含109位不同口音的说话者，每位说话者约400个句子，录制时使用了8种不同的麦克风，并在两种背景噪声条件下进行。