Sounds of Home: A Speech-Removed Residential Audio Dataset
收藏资源简介:
Sounds of Home数据集是由萨里大学的视觉、语音和信号处理中心创建,旨在支持智能家庭应用中的声音事件检测研究,特别是针对老年人的福祉。该数据集包含1344个音频文件,总计1342小时,记录了8名55-80岁参与者在家中7天的声音环境。数据集通过部署音频记录系统并开发自动语音移除管道来保护隐私,确保音频记录中不包含可识别的语音信息。该数据集适用于训练和基准测试声音事件检测模型,特别适用于家庭环境中的应用,旨在通过声音监测技术提升老年人的生活质量。
The Sounds of Home dataset was created by the Centre for Vision, Speech and Signal Processing at the University of Surrey. It is designed to support research on sound event detection in smart home applications, with a specific focus on the well-being of older adults. This dataset comprises 1,344 audio files totaling 1,342 hours, capturing the acoustic environments of 8 participants aged 55 to 80 over a 7-day period in their own homes. To safeguard privacy, the dataset was developed using deployed audio recording systems and an automatic speech removal pipeline, ensuring that no identifiable speech content is present in the audio recordings. This dataset is applicable for training and benchmarking sound event detection models, particularly for home-environment-based applications, and aims to enhance the quality of life of older adults through sound monitoring technologies.

- 1The Sounds of Home: A Speech-Removed Residential Audio Dataset for Sound Event Detection视觉、语音和信号处理中心,萨里大学,英国 · 2024年



