434小时德语自然对话手机采集语音数据

德语语音数据_对话（手机），基于30余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500余名来自不同地域和文化背景的德国本土人录制，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。

German Spoken Dialogue Dataset (Mobile-recorded). This dataset is simulated and recorded based on more than 30 common topics. It is annotated with various attributes including text transcripts, sentence-level timestamps, speaker identities and genders. Recorded by over 500 native German speakers from diverse regions and cultural backgrounds, the dataset boasts high accuracy. It provides abundant resources for speech recognition-related research and applications, and has been validated by multiple AI companies that it helps models deliver outstanding performance when facing real-world diversities. We strictly abide by data protection laws and privacy regulations to safeguard users' privacy and legitimate rights and interests throughout the entire process of data collection, storage and usage. All data complies with GDPR, CCPA and PIPL.

搜集汇总

背景与挑战

背景概述

该数据集包含434小时的德语自然对话语音，通过手机采集，基于30多个常见主题模拟录制，由500多名德国本土人参与。数据标注了文本、时间戳、说话人身份和性别等属性，适用于语音识别研究和应用，并严格遵循数据保护法规。