200人中文唤醒词手机采集语音数据_朗读【数据堂】

中文唤醒词语音数据_朗读（手机），基于给定的脚本朗读并模拟录制，涵盖多个唤醒词，内容丰富。此数据集标注了文本内容等多种属性，采集自200名中国本地人，每人180句；录制人员男女比例均匀；采集环境多样，覆盖七大方言区。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

Chinese Wake-Up Word Speech Dataset (Reading Style, Mobile Phone Recordings) is developed through script-based reading and simulated recording, covering multiple wake-up words with rich content. This dataset is annotated with text content and various other attributes. It is collected from 200 local Chinese residents, with 180 sentences per participant, and the gender ratio of the recorders is evenly balanced. The collection environments are diverse, covering all seven major Chinese dialect regions. The texts have been manually proofread with high accuracy, providing abundant resources for speech recognition-related research and applications. Verified by multiple AI companies, this dataset helps models deliver excellent performance when facing real-world diversities. We strictly comply with data protection regulations and privacy requirements to safeguard users' privacy and legitimate rights and interests throughout the entire process of data collection, storage and usage. All data conforms to GDPR, CCPA and PIPL.

搜集汇总

背景与挑战

背景概述

该数据集包含200名中国人通过手机朗读中文唤醒词的语音数据，每人录制180句，覆盖多个唤醒词和七大方言区，采用16kHz、16bit的wav格式，标注文本内容准确率达99%，适用于语音识别研究。