King-ASR-384-19_温州重口音普通话识别语音库（手机）

该语音数据集由专业团队严格遵循标准化流程定向采集，语音形态丰富、字符准确率高，可支撑语音识别、跨文化交流等相关研究。数据集遵循标准化管理规范，保障数据的完整性、一致性与可用性，适配具身智能应用、学术研究等多类场景，可灵活满足科研机构、企业等不同使用者需求，为各行业高质量发展提供可靠数据支撑，充分发挥数据赋能实效，助力推动人工智能技术落地迭代，加速全球数字化与智能化转型进程。

This speech dataset was directionally collected by a professional team in strict compliance with standardized procedures, boasting rich speech varieties and high character accuracy, and can support relevant research including speech recognition and cross-cultural communication. The dataset follows standardized management norms to guarantee the integrity, consistency and availability of the data, and is suitable for multiple scenarios such as embodied intelligence applications and academic research. It can flexibly meet the demands of different users including scientific research institutions and enterprises, providing reliable data support for the high-quality development of various industries, fully exerting the practical empowering effect of data, facilitating the implementation and iteration of artificial intelligence technologies, and accelerating the global digital and intelligent transformation process.

搜集汇总

背景与挑战

背景概述

该数据集是一个针对温州地区重口音普通话的语音识别库，由专业团队采用标准化流程通过手机采集，具备语音形态丰富和字符准确率高的特点。它遵循严格的管理规范，保障数据的完整性、一致性与可用性，适用于语音识别、跨文化交流等研究，并适配具身智能应用、学术研究等多类场景，服务于科研机构和企业用户。数据集规模为161 GB，模态为音频/语音识别，分类为行业通识，需付费授权使用。