The data consists of multiple sets of products, each with a different type of microphone arrays. Noise data is collected from real home scenes of the indoor residence of ordinary residents. The data set can be used for tasks such as voice enhancement and automatic speech recognition in a home scene.
For more details, please refer to the link: https://www.nexdata.ai/datasets/255?source=Github
Microphone array 48kHz 32bit(Floating point)
noises of TV,refrigerator,air-conditioner,kitchen ventilator,washing machine,airpurifier,and dust collector;multi-people dialogue
about 620 pieces of speec hdata(with window open/closed),and the time of each piece of data is around 1 min
living room, kitchen, bedroom, with window open/closed
MEMS microphone array
8-microphone ring array,6(6+1)-microphone ring array,4(4+1)-microphone ring array,4-microphone liner array,3(3+1)-microphone ring array,2-microphone liner array
speech enhancement under household scene,and speech recognition
Commercial License