数据堂—234小时日语手机采集语音数据_朗读

日本日语语音数据_朗读（手机），基于给定的脚本朗读并模拟录制，涵盖21万句常用日语书面语及常用日语口语。此数据集标注了文本内容、时间戳等多种属性，由799名来自不同地域和文化背景的日本本土人在安静的室内、街道、餐馆等环境下录制，口音正宗，文本经过人工校对，准确率高，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。

Japanese Reading Speech Dataset (Mobile Collected): This dataset is constructed by having speakers read aloud from pre-defined scripts and conduct simulated recordings, covering 210,000 sentences of common written and spoken Japanese. It is annotated with multiple attributes including text content and timestamps. Recorded by 799 native Japanese speakers from diverse regions and cultural backgrounds across various environments such as quiet indoor spaces, streets, and restaurants, the dataset features authentic accents. All texts have undergone manual proofreading to ensure high accuracy. This dataset provides rich resources for speech recognition-related research and applications, and has been verified by multiple AI companies that it helps models achieve excellent performance when facing real-world diversity. We strictly abide by data protection regulations and privacy rules, ensuring the protection of user 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234.8小时的日语朗读语音数据，由799名母语者使用手机在多种环境下录制，覆盖21万句常用日语文本，并标注了文本内容、时间戳等属性。数据格式为16kHz、16bit的WAV文件，适用于语音识别研究与应用测试，且遵循相关隐私法规。