Department of Computing and Informatics- Bournemouth Univerity – Bournemouth, United Kingdom - Jun 2020
- Camargo Saray
- Possamai Marco G
- Tamimi Hossein
Abstract
Many lives are lost every day due to extreme depression. According to the World Health Organisation, 800,000 lives are lost per year due to suicide (Organization, 2020). The lives that are lost to suicide do not have to be vulnerable people that are in unfortunate situations, Suicide can be a terrible thought that can be present in any person's mind regardless of gender, sexuality,ethnicity and appearance. We are proposing a system that possesses the ability to detect a user’s face, analyse the user’s facial expression which will then be classified into emotion and output a calculation that classifies if a user is depressed by training our model using the Fer(2013) dataset (Verma, 2018). However, this alone is not enough to calculate a user’s risk of suicide. This application will impact the medical industry by identifying people who are at risk of suicide by identifying people who are displaying signs of depression which could ultimately lead to suicide.
Dataset:
The Facial Expression Recognition 2013 (FER-2013) (Verma, 2018) that will be used to train the model. This dataset consists of multiple emotion categories which contain images of different people portraying the relevant emotion. These categories will consist of anger, disgust, fear, happiness, sadness, surprise and neutrality. The portraits of these images will be a 48x48 pixel values. https://www.kaggle.com/deadskull7/fer2013
Further details of this project are found on the Report_Suicide_Risk_Recognition_Model.pdf file