Facial Keypoint Detection

Facial Keypoint Detection is one of the most challenging and important topics, which is taken into account in realm of Machine Learning and Computer Vision. The importance originates in its applications, the most three important of which are: 1Face recognition, which is of high importance in identification as an example. 2Medical purposes and Biomedical applications, Medical surgeries and psychiatric tests like analyzing facial expressions could be carried out on patients using Facial Detection. 3Tracking Faces in Images and Videos, these days many applications have a built-in feature of facial analysis. In this report, first, we briefly introduce the topic in section I. Then we proceed with the explanation of the methods and algorithms we used in section II including our proposed method PCLWLR. Then in section III, we mathematically formulate the PCLWLR, and finally, we report the achieved RMSE values for different methods in addition to providing plots illustrating keypoints detected on images. We conclude in section V and summarize the results. Keywords— Keypoint; target variables; edge detection; PCA; PCLWLR; LWLR