127小时巴西葡萄牙语语音数据_对话（手机）【数据堂】

巴西葡萄牙语语音数据_对话（手机），基于约三十个常见主题来模拟录制。此数据集标注了文本内容、句时间戳、说话人身份、性别等多种属性，由142名来自不同地域和文化背景的巴西本土人录制，准确性高，为语音识别相关研究及应用提供了丰富的资源，经多家AI公司验证：有助于模型面对真实世界的多样性时能够表现出色。我们严格遵循数据保护法规和隐私规定，确保数据采集、存储和使用的过程中维护用户的隐私和合法权益，所有数据均遵循GDPR, CCPA, PIPL

Brazilian Portuguese conversational speech dataset (mobile-recorded), which is simulated and recorded based on approximately thirty common topics. This dataset is annotated with multiple attributes including text transcripts, sentence-level timestamps, speaker identities, genders and other relevant metadata. Recorded by 142 Brazilian native speakers from diverse regional and cultural backgrounds, the dataset boasts high accuracy and provides abundant resources for speech recognition-related research and applications. Verified by multiple AI companies, it enables models to achieve excellent performance when coping with the diversity of real-world scenarios. We strictly comply with data protection laws and privacy regulations to safeguard users' privacy and legitimate rights and interests throughout the entire process of data collection, storage and utilization. All data in this dataset adheres to GDPR, CCPA and PIPL.

搜集汇总

背景与挑战

背景概述

该数据集包含127小时的巴西葡萄牙语对话语音，由142名巴西本土人士基于约30个常见主题录制，格式为16kHz、16bit的未压缩WAV文件，并标注了文本、时间戳、说话人身份和性别等属性，词准确率达98%，专为语音识别研究设计，仅对企业合作购买开放。