FACE RECOGNITION WITH BILINEAR CNNS: A REVIEW

Abstract

The identification of faces from real data, captured images, sensor photos, and database 
images is a challenging endeavor. This is because there is a wide range of facial appearances, the 
illumination effect, and the intricacy of the image backdrop.  The use of image processing and biometric 
technology for facial recognition is one of the most effective and significant applications of these 
technologies.  This research presents a discussion of face identification methods and algorithms that have 
been developed by a number of researchers using artificial neural networks (ANN). These algorithms 
have been utilized in applications that are not related to image processing and pattern recognition.  This 
article will also discuss how artificial neural networks (ANN) will be utilized for facial recognition, as 
well as how it is superior to other approaches in terms of efficiency.  In order to present an overview of 
face recognition using ANN, there are many different ways that are provided by ANN.  As a consequence 
of this, this research includes an exhaustive analysis of face recognition experiments and systems that 
make use of a wide variety of artificial neural network (ANN) methodologies and algorithms.  Through 
the course of this study effort, both the positive and negative aspects of the aforementioned literature 
studies and systems were incorporated, in addition to a performance examination of a number of different 
ANN techniques and algorithms.  During the past twenty years, the discipline of shape detection has 
emerged as the most exciting sector of the scientific community.  Object identification is the focus of our 
discussion as we provide an innovative approach to identifying shapes that are similar to one another.  The 
process of determining the correspondences between the vertices of two shapes and then using those 
correspondences to estimate an aligning transform prior to determining the degree of similarity between 
the shapes.  In this article, we will go over some of the most important aspects of face detection, which are 
useful in a wide range of applications. These applications include face recognition, image classification, 
the ability to track faces, facial feature extraction, person identification, identification number, record 
keeping and access control, grouping, biometric scientific knowledge, human computer interaction (HCI) 
system, electronic beauty products, and many more.  Prior to that, I would like to go over some 
well-known facial recognition techniques and then some image processing approaches. This is because 
we won't be able to effectively identify someone unless we remove the primary parts of their face, which 
are their eyes, nose, and mouth.  Calculating the difference between the two shapes involves determining 
the percentage of the matching mistake that exists between locations that are comparable.  As a method 
for classifying recognition, we make use of a closest neighbor classification. 

,