303小时中国中英混读语音数据_朗读（手机）【数据堂】

中国中英混读语音数据_朗读（手机），基于给定的脚本朗读并模拟录制，内容覆盖通用口语、人机交互领域。此数据集标注了文本内容等多种属性，由1,113名来自不同地域和文化背景的人员录制，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

Chinese-English Code-switching Speech Dataset (Read Speech, Mobile Recording): Recorded by participants reading given scripts in simulated mobile recording scenarios, with content covering general spoken language and human-computer interaction domains. This dataset is annotated with multiple attributes including text content, and was recorded by 1,113 participants from diverse regions and cultural backgrounds with high annotation accuracy. It provides rich resources for research and applications related to speech recognition, and has been verified by multiple AI companies that it enables models to perform excellently when facing real-world diversity. We strictly comply with data protection regulations and privacy rules to safeguard users' privacy and legitimate rights and interests throughout the data collection, storage and use processes, and all data 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303小时的中国中英混读语音数据，基于脚本朗读录制，覆盖通用口语和人机交互领域，由1,113名中国人员参与录制，语言包括中文普通话和英文。数据格式为16kHz、16bit的wav音频，标注了文本内容，句准确率达97%，经多家AI公司验证有助于提升模型在真实场景中的表现。该数据为商业数据，仅面向企业合作购买。