数据堂—338小时西班牙语手机采集语音数据

338小时西班牙语手机采集语音数据，由800名西班牙、墨西哥、阿根廷等国的西班牙语母语人员参与录制；录音环境安静；所有文本由人工转写，句准确率95%。338小时西班牙语手机采集语音数据可应用于语音识别、机器翻译、声纹识别等

This 338-hour Spanish speech dataset was collected via mobile phones, recorded by 800 native Spanish speakers from countries including Spain, Mexico, Argentina and other regions. All recordings were made in quiet environments, and all transcriptions were manually completed with a sentence-level accuracy of 95%. This dataset can be applied to tasks such as speech recognition, machine translation and speaker recognition.

搜集汇总

背景与挑战

背景概述

该数据集包含338小时通过手机采集的西班牙语语音，由800名来自西班牙、墨西哥等国家的母语者录制，内容涵盖新闻、评论等多种类型。数据适用于西班牙语语音识别模型的评估任务，也可用于机器翻译和说话人验证等应用。