VAHA (POLYPHONE II)

<h3>Introduction</h3> <p>Voice Across Hispanic America (VAHA) is a corpus of Spanish telephone speech, recorded digitally from 915 native speakers of Spanish in various parts of the United States. With nearly 39,000 recorded and transcribed utterances, VAHA will be useful for a variety of research studies, but it is intended primarily for speech technology research and development in telecommunications applications. It is patterned after Macrophone (1), an American English corpus (<a href="http://catalog.ldc.upenn.edu/LDC94S21" rel="nofollow">LDC94S21</a>) which is widely used for this purpose. </p><h3>Data</h3> <p>This corpus was collected by Texas Instruments in Dallas, TX for the Linguistic Data Consortium. </p><h3>Updates</h3> <p>There are no updates at this time.</p> </br> Portions © 1996 Trustees of the University of Pennsylvania<h3>简介</h3> <p>跨拉美西班牙语区语音库（Voice Across Hispanic America，简称VAHA）是一套西班牙语电话语音语料库，数据采集自美国各地的915名西班牙语母语使用者，全部采用数字方式录制。该语料库包含近3.9万条经录制与转写的语音话语，可支撑多种研究工作，但其核心应用场景为电信领域的语音技术研发。本语料库的构建参考了Macrophone（参考文献[1]）——一款被广泛应用于同类研究场景的美式英语语料库（<a href="http://catalog.ldc.upenn.edu/LDC94S21" rel="nofollow">LDC94S21</a>）。</p><h3>数据</h3> <p>本语料库由位于美国德克萨斯州达拉斯市的德州仪器（Texas Instruments）受语言学数据联盟（Linguistic Data Consortium）委托采集完成。</p><h3>更新情况</h3> <p>目前暂无更新计划。</p> </br> Portions © 1996 Trustees of the University of Pennsylvania

搜集汇总

背景与挑战

背景概述

VAHA (POLYPHONE II) 是一个1996年发布的西班牙语电话语音数据集，包含915名母语者的近39,000条录音和转录话语，采样率为8000 Hz。该数据集专为电信语音识别技术研发设计，基于英语语料库Macrophone的模式构建，适用于语音技术研究和应用开发。