Automatic Detection of Chewing and Swallowing Using Multichannel Sound Information
Automatic Detection of Chewing and Swallowing Using Multichannel Sound Information
复制标题
DOI：
10.1109/lifetech52111.2021.9391873
复制
发表时间：
2021-03
期刊：
影响因子：
--
通讯作者：
Akihiro Nakamura;Takato Saito;Daizo Ikeda;Ken Ohta;H. Mineno;M. Nishimura
中科院分区：
文献类型：
--
作者：
Akihiro Nakamura;Takato Saito;Daizo Ikeda;Ken Ohta;H. Mineno;M. Nishimura
监控饮食行为对保持健康很重要。因此，必须开发一种能够自动检测咀嚼和吞咽并提供诸如咀嚼位置等方面的信息的技术。本文特别关注进食时嘴里发出的声音，并演示了一种使用皮肤接触式麦克风自动提取所需信息的方法。除了传统的放置在耳朵下方的左右声道喉咙麦克风外，该技术还利用从上下两声道喉咙麦克风和连接到颈部的接触式麦克风获得的声音相关信息来稳定和改进吞咽的检测。
It is important to monitor eating behavior to maintain health. A technique that can automatically detect chewing and swallowing and provide information on aspects such as chewing position must thus be developed. This paper focuses especially on the sounds made in the mouth while eating and demonstrates a method that uses skin contact microphones to automatically extract the desired information. In addition to conventional left and right channel throat microphones placed under the ear, the proposed technique utilizes sound-related information obtained from upper and lower 2-channel throat microphones and a contact microphone attached to the neck to stabilize and improve the detection of swallowing.