Department of Computer Engineering
CS 590 SEMINAR
End-to-End Spatio-Temporal Pain Intensity Estimation
Diyala Nabeel Ata Erekat
Computer Engineering Department
Pain is an uncomfortable feeling that makes pain assessment extremely crucial for proper patient recovery and good pain control, however the standard clinical assessment of pain is limited primarily to self reported pain (e.g., using Visual Analog Scale (VAS), Sensory Scale (SEN), Affective Motivational Scale (AFF)). Automatic facial expression analysis has emerged as a potential solution for objective, reliable, and valid pain measurement. While previous research on automatic pain estimation has focused primarily on investigating spatial information in static images to estimate frame-level pain using Prkachin and Solomon Pain Intensity (PSPI) metric, we propose an end-to-end spatio-temporal regression framework using a modified loss to estimate 4 different pain scores; Observer Pain Intensity (OPI) and 3 different self-reported; SEN, AFF and VAS, for a video sequence.
DATE: 02 December 2019, Monday @ 16:10