Newsroom

Canavan and Uddin winners in IEEE FG 2020 EmoPain Challenge

December 3, 2020

CSE Assistant Professor Shaun Canavan and PhD student MD Taufeeq Uddin won two tracks in the FG 2020 EmoPain Challenge. 

The goal of the EmoPain challenge was to develop machine learning models to assess the pain levels and relevant protective behaviors, such as guarding, of patients with chronic lower back pain. Canavan and Uddin competed in two out of the three tracks in the challenge and won both of these tracks. The model they developed was a multimodal sequential machine learning model that takes into account human physiological and body movement data, context (such as physical activities performed by the patients), and relevant demographic/meta information. The proposed and developed model was assessed by the benchmark test dataset by the University College London and the University of Nottingham, UK. The developed model demonstrated that for a fair assessment of pain, both human physiology information and context in which physiological information was captured were essential for both high recognition performance and sound modeling. The developed model has the potential to improve the self-care of patients with lower back pain and to augment the productivity of healthcare professionals (such as doctors, nurses) and caregivers.