Automated face analysis by feature point tracking has high concurrent validity with manual FACS coding

The face is a rich source of information about human behavior. Available me thods for coding facial displays, however, are human-observer dependent, la bor intensive, and difficult to standardize. To enable rigorous and efficie nt quantitative measurement of facial displays, we have developed an automa ted method of facial display analysis. In this report, we compare the resul ts with this automated system with those of manual FAGS (Facial Action Codi ng System, Ekman & Friesen, 1978a) coding. One hundred university students were videotaped while performing a series of facial displays. The image seq uences were coded from videotape by certified FAGS coders. Fifteen action u nits and action unit combinations that occurred a minimum of 25 times were selected for automated analysis. Facial features were automatically tracked in digitized image sequences using a hierarchical algorithm for estimating optical flow. The measurements were normalized for variation in position, orientation, and scale. The image sequences were randomly divided into a tr aining set and a cross:validation set, and discriminant function analyses w ere conducted on the feature point measurements. In the training set, avera ge agreement with manual FAGS coding was 92% or higher for action units in the brow, eye, and mouth regions. In the cross-validation set, average agre ement was 91%, 88%, and 81% for action units in the brow, eye, and mouth re gions, respectively. Automated face analysis by feature point tracking demo nstrated high concurrent validity with manual FAGS coding.