数据堂300小时印度英语自然对话手机采集语音数据

印度英语自然对话手机采集语音数据，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390名来自不同地域和文化背景的印度本土人录制，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。

This is a mobile-collected Indian English natural conversational speech dataset, simulated and recorded based on over 30 common topics. It is annotated with multiple attributes including text transcripts, sentence-level timestamps, speaker identities, gender, and other relevant information. Recorded by 390 native Indian speakers from diverse regional and cultural backgrounds, it features high annotation accuracy. This dataset provides abundant resources for research and applications related to speech recognition. Verified by multiple AI enterprises, it enables models to perform exceptionally well when confronted with real-world diversity. We strictly abide by data protection laws and privacy regulations to protect users' privacy and legitimate rights and interests throughout the entire process of data collection, storage and utilization. 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300小时的印度英语自然对话语音数据，由390名母语者通过手机在安静室内环境下录制，覆盖30多个常见主题。数据以16kHz、16bit单声道WAV格式提供，标注包括文本内容、时间戳、说话人身份和性别，单词准确率达到98%。