A dataset of genuine and synthetic Portuguese speech covering European, Brazilian, African, and regional accents - An open resource for speech deepfake detection and trustworthy AI research

IF 1.9 Q3 MULTIDISCIPLINARY SCIENCES

Data in Brief Pub Date : 2026-08-01 Epub Date: 2026-08-30 DOI:10.1016/j.dib.2026.113214

André Miguel Viegas Oliveira, Rafael Geraldo dos Santos, Helder Daniel, José Valente de Oliveira

{"title":"A dataset of genuine and synthetic Portuguese speech covering European, Brazilian, African, and regional accents - An open resource for speech deepfake detection and trustworthy AI research","authors":"André Miguel Viegas Oliveira, Rafael Geraldo dos Santos, Helder Daniel, José Valente de Oliveira","doi":"10.1016/j.dib.2026.113214","DOIUrl":null,"url":null,"abstract":"<div><div>Portuguese is spoken by approximately 300 million people, making it one of the most widely spoken languages in the world, with speakers distributed across every continent. Despite its global reach, Portuguese remains significantly underrepresented in current synthetic speech detection technologies. This dataset aims to address this critical gap in the existing detection solutions. This dataset may be used in tasks related to biometric security, speaker verification, anti-spoofing systems, and deepfake protection, supporting the development and evaluation of models for Portuguese language speech data.</div><div>The Synthetic Speech Detection - Portuguese (SSD_PT) dataset consists of 242,156 audio samples in the Portuguese language, containing short speech segments ranging from 1 to 7 s each. The dataset is organized into two balanced classes: genuine speech samples (bonafide) and synthetic speech samples (spoof).</div><div>For the Brazilian Portuguese portion of the dataset, the genuine speech samples were collected from publicly available repositories, namely VoxForge, Mozilla Common Voice, and OpenSLR. As no publicly available repositories with comparable coverage were found for the remaining Portuguese varieties, the genuine speech samples for the remaining portions of the dataset were collected from television interviews, radio programs, and podcasts representing different regions of the Lusophone world, including mainland Portugal, the Azores, Madeira, Angola, and Mozambique.</div><div>The synthetic speech samples were generated from genuine speech material. For Brazilian Portuguese, a text-to-speech model based on Coqui TTS (XTTS-v2) was used. For other Portuguese variants, voice conversion techniques were applied using the RVC (Retrieval-based Voice Conversion) framework, requiring the creation of speaker-specific models trained on multiple audio segments to capture individual vocal characteristics.</div><div>All audio files were standardized to a 16 kHz sampling rate and converted to mono format. The dataset is split into training, development, and evaluation subsets, accompanied by protocol files that associate each sample with its corresponding label.</div></div>","PeriodicalId":10973,"journal":{"name":"Data in Brief","volume":"68 ","pages":"Article 113214"},"PeriodicalIF":1.9000,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Data in Brief","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2352340926007614","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2026/8/30 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"MULTIDISCIPLINARY SCIENCES","Score":null,"Total":0}

引用次数: 0

Abstract

Portuguese is spoken by approximately 300 million people, making it one of the most widely spoken languages in the world, with speakers distributed across every continent. Despite its global reach, Portuguese remains significantly underrepresented in current synthetic speech detection technologies. This dataset aims to address this critical gap in the existing detection solutions. This dataset may be used in tasks related to biometric security, speaker verification, anti-spoofing systems, and deepfake protection, supporting the development and evaluation of models for Portuguese language speech data.

The Synthetic Speech Detection - Portuguese (SSD_PT) dataset consists of 242,156 audio samples in the Portuguese language, containing short speech segments ranging from 1 to 7 s each. The dataset is organized into two balanced classes: genuine speech samples (bonafide) and synthetic speech samples (spoof).

For the Brazilian Portuguese portion of the dataset, the genuine speech samples were collected from publicly available repositories, namely VoxForge, Mozilla Common Voice, and OpenSLR. As no publicly available repositories with comparable coverage were found for the remaining Portuguese varieties, the genuine speech samples for the remaining portions of the dataset were collected from television interviews, radio programs, and podcasts representing different regions of the Lusophone world, including mainland Portugal, the Azores, Madeira, Angola, and Mozambique.

The synthetic speech samples were generated from genuine speech material. For Brazilian Portuguese, a text-to-speech model based on Coqui TTS (XTTS-v2) was used. For other Portuguese variants, voice conversion techniques were applied using the RVC (Retrieval-based Voice Conversion) framework, requiring the creation of speaker-specific models trained on multiple audio segments to capture individual vocal characteristics.

All audio files were standardized to a 16 kHz sampling rate and converted to mono format. The dataset is split into training, development, and evaluation subsets, accompanied by protocol files that associate each sample with its corresponding label.

查看原文本刊更多论文

真实和合成的葡萄牙语语音数据集，涵盖欧洲、巴西、非洲和地区口音——语音深度假检测和可靠的人工智能研究的开放资源

大约有3亿人使用葡萄牙语，使其成为世界上使用最广泛的语言之一，使用者遍布各大洲。尽管葡萄牙语遍及全球，但在当前的合成语音检测技术中，葡萄牙语的代表性仍然明显不足。该数据集旨在解决现有检测解决方案中的这一关键差距。该数据集可用于与生物识别安全、说话人验证、反欺骗系统和深度假冒保护相关的任务，支持葡萄牙语语音数据模型的开发和评估。合成语音检测-葡萄牙语（SSD_PT）数据集由242156个葡萄牙语音频样本组成，每个样本包含1到7秒的短语音片段。数据集被组织成两个平衡的类：真实语音样本（bonafide）和合成语音样本（spoof）。对于数据集的巴西葡萄牙语部分，真实的语音样本是从公开可用的存储库中收集的，即VoxForge， Mozilla Common Voice和OpenSLR。由于没有找到具有可比覆盖范围的公共可用存储库，因此数据集其余部分的真实语音样本是从代表葡语世界不同地区的电视采访、广播节目和播客中收集的，包括葡萄牙大陆、亚速尔群岛、马德拉群岛、安哥拉和莫桑比克。合成语音样本由真实语音材料生成。对于巴西葡萄牙语，使用基于Coqui TTS （XTTS-v2）的文本到语音模型。对于其他葡萄牙语变体，使用RVC（基于检索的语音转换）框架应用语音转换技术，需要创建针对多个音频片段进行训练的说话人特定模型，以捕获个人语音特征。所有的音频文件被标准化到16千赫采样率和转换为单声道格式。数据集被分成训练、开发和评估子集，并伴随着将每个样本与其相应标签关联起来的协议文件。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Data in Brief MULTIDISCIPLINARY SCIENCES-

CiteScore

3.10

自引率

0.00%

发文量

996

审稿时长

70 days

期刊介绍： Data in Brief provides a way for researchers to easily share and reuse each other''s datasets by publishing data articles that: -Thoroughly describe your data, facilitating reproducibility. -Make your data, which is often buried in supplementary material, easier to find. -Increase traffic towards associated research articles and data, leading to more citations. -Open up doors for new collaborations. Because you never know what data will be useful to someone else, Data in Brief welcomes submissions that describe data from all research areas.