数据堂—1,030小时上海方言手机采集语音数据

1,030小时上海方言手机采集语音数据，由2,956名上海方言发音人，在安静的室内环境下的录音数据。录音内容广泛，包括多领域客户咨询、短信、数字、上海POI等。1,030小时上海方言手机采集语音数据的语料无重复，平均句长12.68字。匹配主流安卓、苹果系统手机

1,030 hours of Shanghai dialect speech data collected via mobile phones, from 2,956 Shanghai dialect speakers, with all recordings performed in quiet indoor environments. The recorded content spans diverse domains, including multi-category customer consultations, short text messages, numerals, and Shanghai Points of Interest (POIs). This 1,030-hour Shanghai dialect speech corpus has no duplicate samples, with an average sentence length of 12.68 Chinese characters. It is compatible with mainstream Android and Apple iOS mobile devices.

搜集汇总

背景与挑战

背景概述

该数据集包含1,030小时的上海方言手机采集语音数据，专用于中文语音识别模型的测试任务。数据由2,956名使用者在安静室内录制，内容覆盖多领域查询、短信、数字和上海地标，语料无重复，平均句子长度为12.68字符，并兼容主流手机系统。