Dauruxu : detección de emociones de personas y sus actividades para el apoyo en la evaluación de factores de riesgo psicosocial

Psychosocial risk assessment has played a dominant role in ensuring the well-being and health of people. However, mechanisms such as interviews and questionnaires are susceptible to obtaining biased results due to the lack of data that cannot be acquired during evaluations. This work proposes an architecture to identify activities and emotions implicit in current questionnaires and that have the potential to be detected by cameras. Through computer vision, features are extracted from the video frames which are used as predictors for classification tasks. The quantification of indicators based on the detection of activities and emotions will provide additional data to support psychosocial risk assessments.