Research with wearable sensor

This study measured physical activity during remote work and explored the movements associated with performance. During remote working, we almost sit in front of a computer. Because the movements are slight, a new neural net should be created to distinguish subtle chest movements. Study 1 classified chest movements into six categories: walking, rising and sitting, standing still, rotating, swaying, and rocking. (n = 8, ages 19–62, 6 males, 2 females) The LSTM model showed a better correct response rate (M = 83.8%) than the CNN model. Study 2 showed that chest swing affects reaction time and that reaction time affects work accuracy. <br>