Week 1 Journal

This week, our topic was confirmed and we conducted a literature review. Jennifer and I chose to detect heart rate through videos of people's faces. We read a few papers, and found that the way most people performed this project was by taking a video input, pre-processing the frames so that the face is detected, and separating each frame into three channels - R, G, B. In each channel, the intensity of the pixels are compared throughout the frames. Using the change in intensity and time, a heart rate is predicted at the end. Dr. Briggs and Tianyi approved of our pitch, given that the input of our project would not require an expensive camera to record faces. However, comparing the papers we read, a video input using a cellphone worked fine. As long as the LCDK can process mp4 files, we will be using our phones to record videos of faces.