Mental-state estimation model with time-series environmental data regarding cognitive function

Abstract

It is expected that understanding and estimating the human mental state will be helpful for mental health measures, improving learning and labor work efficiency, and preventing human error. Our research group has been developing a methodology and model for estimating the mental state of humans in their environment from indoor environmental data regarding cognitive function obtained by wireless sensor network technology. This study constructed a model to estimate mental states from multidimensional time-series indoor environmental data such as temperature, humidity, and illumination. The experiment results show that the proposed system shows higher estimation accuracy. In particular, the CNN-LSTM model with multidimensional time-series indoor environmental data shows about 90% estimation accuracy. The results obtained in this study, showing that mental states can be determined with high accuracy from environmental data, are helpful for future research approaches. It may also contribute to the creation of a less stressful environment.