Almost 800000 people die to suicide around the world per year. Indonesia is one of the country listed in a medium to high risk of suicide rate. Depression is the most common psychiatric disorder in people who die by suicide. It is hard to be recognised as depressed person do not show any symptoms of depression. Early detection is hence required to screen the symptoms. Some methods to screen the symptoms are by using technology. One of the state-of-the-art of the research is by implementing virtual humans to help psychologist to automatically detect the symptoms of depression. The works of virtual humans as mental health tools have been quite popular among the researchers in the UK and USA. However, no similar research has been done in Indonesia. We postulate that using virtual human can help psychologist to automatically detect the symptoms of depression as well as to monitor the patients. This also will deal with the loneliness of patients by having a friend to talk. First, we will collect audio-visual datasets to be trained into models using deep learning. The models then will be implemented to a virtual human and finally the system will be evaluated with the professional psychologists.

Mental Health, Emotion Recognition, Chatbot, Facial Expression Recognition, Natural Language Processing

• Dr. Esther Widhi Andangsari, M.Si., Psi.

• Ir. Andry Chowanda, S.Kom., MM, Ph.D., MBCS, CCP, CME, IPM, SMIEEE

• Violitta Yesmaya, S.Kom., M.T.I.