数据堂—101小时录音笔采集场景噪音数据

录音笔采集场景噪音数据为多场景噪音数据，涵盖地铁、超市、餐厅、马路、机场、展厅、高铁、高速公路、市内道路、电影院等多个日常生活场景。该数据均使用专业录音笔Sony ICD-UX560F录制，采用高采样率、双声道格式收集，录音清晰自然。录音笔采集场景噪音数据的有效数据时长101小时

The scene noise dataset collected by digital voice recorders is a multi-scene noise corpus covering various daily life scenarios including subway, supermarket, restaurant, road, airport, exhibition hall, high-speed rail, expressway, urban road, cinema and others. All recordings were made using a professional Sony ICD-UX560F digital voice recorder, collected with high sampling rate and dual-channel format, featuring clear and natural audio quality. The total effective duration of this scene noise dataset is 101 hours.

搜集汇总

背景与挑战

背景概述

该数据集包含101小时由索尼ICD-UX560F录音笔采集的多场景噪音数据，覆盖地铁、超市、街道等日常生活环境，采用44.1kHz双声道WAV格式，适用于中文语音识别模型的测试任务。