Study and Analysis of Facial Landmark Detection Techniques

Authors

  • Manisha M. Kasar, S.H. Patil

Abstract

Facial landmark detection and position of individual point on face is an essential procedure for various successive face operations that has scope from biometric recognition to the awareness of psychological states. Facial landmark is to be identified point by point such as eyes, nose, mouth, eyebrows on the face. Facial landmark is a collection of characteristics that can be differentiated in the person face with the open eye. It can be utilized to regulate either person beings present in the image or not, recognize who the human is or identify the location of a face when accurate. The motive of this study is to provide a summary of landmarking techniques and their improvement across the past years, classifies them and exhibit comparative performance of the landmark techniques. Techniques for identifying facial feature points can be categorized into two types. First type is depending on conventional image processing methods. Another type is depending on machine learning technique where feature points is recognized through training facial features. The correctness of facial feature point localization is associated with various emotions and is influenced by the outcome of face detection. Various applications use face position or posture detection in upcoming research area. This study utilizes various methods which differentiates real world set of facial feature points. We study the primary movements and specify present limitation with the assumption that. This study will give additional motivation for the frequently required high-performance and actual face landmarking worked at database images.

Published

2020-12-01

Issue

Section

Articles