数据堂—500小时越南语自然对话手机采集语音数据

500小时越南语自然对话手机采集语音数据由约750名越南本土人参与录制，录音人男女比例均衡，500小时越南语自然对话手机采集语音数据无预设语料，为确保对话的流畅自然，由录音人根据自己熟悉的话题展开对话并录制

This 500-hour Vietnamese natural conversational speech dataset was collected via mobile phones, with approximately 750 Vietnamese native speakers participating in the recording. The gender ratio of the participating speakers is evenly balanced between males and females. The dataset contains no pre-scripted corpus; to ensure the naturalness and fluency of the dialogues, the speakers were invited to initiate and record conversations based on topics they are familiar with.

搜集汇总

背景与挑战

背景概述

该数据集包含500小时的越南语自然对话语音数据，由约750名母语者通过手机采集，对话基于熟悉话题进行，无预设脚本，音频格式为16kHz/16bit单声道WAV。它主要用于语音识别模型的测试任务，版权归数据堂所有，采用Apache 2.0许可证。