FMCW Radar-Enabled Eyeblink Detection: An AVMD-Based Solution

Eyeblink detection and analysis are essential in various applications, including human-computer interaction, driver fatigue and drowsiness monitoring, and medical diagnostics. This study proposes a non-invasive technique for eyeblink detection using frequency-modulated continuous-wave (FMCW) radar and adaptive variational mode decomposition (AVMD). Our methodology relies on investigating the impact of subtle motions such as cardiopulmonary activities and head movements by designing a set of targeted experiments and proposing an AVMD-based signal processing solution to mitigate the impact of these interferences and improve blink signal detectability. The experimental results showcase the efficacy and robustness of the proposed method, attaining an F1-score of 93.8% for distances between 25 and 100 cm. These findings emphasize the potential of FMCW radar and AVMD in advanced eyeblink detection and evaluation.