Gender Classification Using Under Floor Vibration Measurements

The ability to automatically classify the gender of occupants in a building has far-reaching applications in multiple areas spanning security and threat detection, retail sales, and possibly biometric identification in smart buildings. While other classification techniques provide potential for gender classification, they face varied limitations such as invasion of privacy, occupant compliance, line of sight, and high sensor density. High-sensitivity accelerometers mounted under the floors provide a robust alternative for occupant classification. The authors take advantage of the highly-instrumented Goodwin Hall on the Virginia Tech campus to measure vibrations of the walking surface caused by individual walkers. A machine learning technique known as Support Vector Machines (SVMs) is used to classify gender. In this study, the gait (i.e. walking) of 15 individual walkers (eight male and seven female) was recorded as they, alone, walked down the instrumented hallway, in multiple trials. The trials were recorded via 14 accelerometers which were mounted underneath the walking surface with the placement of the sensors unknown to the walker. A tenfold-cross-validation method is used to comment on the validity of the algorithm's ability to generalize to new walkers. This work demonstrates that a gender classification accuracy of 88 % is achievable using the underfloor vibration data from the Virginia Tech Goodwin Hall applying an SVM approach.