Non-contact Multi-target Vocal Folds Vibration Detection based on MIMO FMCW Radar

Abstract

This paper presents a non-contact vocal folds vibration detection method for multiple targets based on an integrated 77-GHz multiple-input multiple-output (MIMO) frequency modulated continuous wave (FMCW) radar sensor. The multiple signal classification (MUSIC) based direction-of-arrival (DOA) estimation and linear constraint minimal variance (LCMV) adaptive digital beamforming (ADBF) are applied to obtain the separated signals of multiple subjects. The range-FFT and signal realignment methods are performed on the separated signals to achieve the vocal folds vibration of each subject. Experiments of two subjects speaking simultaneously are carried out. Compared with the reference vibration sensor, the proposed method based on radar can accurately detect the vocal folds vibration of multiple targets at the same time.