数据堂—55小时英国儿童麦克风采集语音数据

英国英语儿童语音数据_朗读（麦克风），基于给定的脚本朗读并模拟录制，由201名英国本地儿童采用高保真麦克风录制，录音内容符合儿童特点，主要为儿童教科书、故事书等内容，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

British English Children's Read Speech Dataset (Microphone Recorded). This dataset was recorded by 201 local British children using high-fidelity microphones, with participants reading aloud based on provided scripts. The recorded content aligns with children's age-appropriate characteristics, mainly consisting of materials from children's textbooks, storybooks and other similar content. The accompanying texts have undergone manual proofreading with high accuracy, providing rich resources for speech recognition-related research and applications. Verified by multiple AI companies, this dataset enables models to perform excellently when confronted with real-world diversity. We strictly comply with data protection regulations and privacy rules, ensuring the protection of users' privacy and legitimate rights and interests throughout the entire process of data collection, storage and utilization. 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55小时由201名英国儿童使用高保真麦克风在安静室内朗读儿童教材和故事书采集的语音数据，专门用于语音识别研究和应用测试。数据以44.1 kHz、16-bit WAV格式提供，句子级转录准确率达95%，遵循Apache 2.0许可证并由数据堂拥有商用版权。