Recognition of Emotional State Based On Handwriting Analysis and Psychological Assessment

Abstract

Emotions describe the physiological states of an individual and are generated subconsciously. They motivate, organize, and guide perception, thought, and action. Emotions can be positive or negative. Negative emotions manifest in the form of depression, anxiety and stress. It is necessary to identify negative emotions of an individual who might be in the need for counseling or psychological treatment. Body signal analysis, handwriting analysis, and psychological assessment are some mechanisms to measure them. In this paper, emotional state is being measured through the person’s handwriting sample analysis and psychological assessment. Psychological assessment is done by using the results of DASS questionnaire attempted by the individual. Convolutional Neural Network (CNN) algorithm is used to find the emotional state of an individual from his/her handwriting sample. Comparative analysis is performed to suggest counseling/medication if required. The final CNN model is formed by using the ensemble method over cross-validation models. The accuracy achieved by the CNN model over the test dataset is 91.25%.