International Journal of Advanced in Management, Technology and Engineering Sciences ISSN NO : 2249-7455
A STUDY OF FACTORS AFFECTING FACE RECOGNITION
Aayushi Bansal
Department of Computer Science & Engineering
Guru Jambeshwar University of Science &Technology, Hissar, Haryana -125001.
Email: [email protected]
Abstract
Face recognition has been one of the most interesting and important topic in research fields as it
deals with security and privacy. The human face plays a critical role in our social association, passing
on individuals' identity. The reasons come from the interest in the design of human-computer
interface and in human visual system on face recognition. These involve knowledge of different
domain and researchers from different disciplines such as neuroscience, psychology, computer vision,
pattern recognition, image processing, and machine learning, etc. In this paper we will study several
factors that result in difficulties of face detection and face recognition.
Keywords: Face Recognition, Face Detection, Feature Extraction.
1. INTRODUCTION
In this paper, we focus on image-based face recognition. Given a picture taken from a digital
camera, we’d like to know if there is any person inside, where his/her face locates at, and who he/she
is. However, we generally separate the face recognition procedure into three steps: Face Detection,
Feature Extraction, and Face Recognition.
INPUT FACE FEATURE FACE
IMAGE DETECTION EXTRACTION RECOGNITION
2. Face Detection
The main function of face detection is to determine (1) whether human faces appear in a
given image or not, and (2) where these faces are located at given image. The expected outputs
of this step are patches containing each face in the input image. Hence to make face recognition
more robust and efficient to design, face alignment are performed to justify the scales and
orientations of these patches. Besides dealing as the pre-processing phase for face recognition,
face detection could be used for region-of-interest detection, retargeting, video and image
classification, etc.
3. Feature Extraction
After the face detection step, human-face patches are extracted from images via feature extraction
phase. If we use these patches for face recognition directly, it may have some disadvantages; first,
each patch usually contains over 1000 pixels, which are too large to build a robust recognition.
Second, face patches may be taken under different illumination, with different face expressions, and
with different camera alignment, and may suffer from occlusion and cluttering background. To
overcome these drawbacks, we perform feature extractions to do dimensionality reduction,
information packing, salience extraction, and noise cleaning. After this step, we transformed face
Volume 8, Issue II, FEB/2018 669 http://ijamtes.org/
International Journal of Advanced in Management, Technology and Engineering Sciences ISSN NO : 2249-7455
patch into a vector with fixed dimension or a set of fiducial points and their corresponding locations.
We can include feature extraction either in face detection or face recognition as per survey form some
literatures.
4. Face Recognition
After formulizing each face patch in the above steps, the last step is to recognize the identities of
these faces. For face recognition, a face database is required to build. For each individual, several
images are taken and their features are extracted and stored in the database. Whenever an input image
came, we perform face detection and face extraction and after this we compare features of each face
with stored database.
We generalized two applications of face recognition: identification and verification. Face
identification means given a face image, we want the system to tell who he / she is and indentify the
individual ; while in face verification, given a face image and on the basis of a guess from the
identification, we want the system to tell true or false about the guess. Face book utilizes facial
acknowledgment programming to help mechanize client labeling in photos. Here's the means by
which facial acknowledgment works in Face book: Each time an individual is tagged in a photo, the
software application stores data about that individual's facial qualities. At the point when enough
information has been gathered about a man to recognize them, the framework utilizes that data to
distinguish a similar face in various photos, and will along these lines recommend labeling those
photos with that individual's name.
5. Factors of human appearance variations
There are several factors that may result in difficulties to perform face detection and face
recognition. Face patches may be taken under different illumination, with different face
expressions, and with different camera alignment, and may suffer from occlusion and cluttering
background. Some of the factors we need to concern are: face pose, face expression,
illumination, RST (rotation, scale, and translation) variation, cluttering, and occlusion.
5.1 Illumination
The appearance of a face will vary drastically when the illumination changes. Variations in
lighting conditions make face recognition an even more challenging and difficult task. The
illumination changes have been discussed in numerous face detection and recognition
researches. This variety is caused by different lighting conditions in which the picture is taken
and is considered to have bigger appearance contrasts than the distinctions caused by various
personalities. Under some enlightenment conditions, we can neither guarantee the recognizable
proof nor precisely call attention to the places of facial highlights.
5.2 Face Pose
The pose variation results from different angles and locations during the face recognition
process. This variation changes the spatial relations among facial highlights and causes serious
distortion on the customary appearance-based face recognition algorithms such as eigenfaces
and fisherfaces.
Volume 8, Issue II, FEB/2018 670 http://ijamtes.org/
International Journal of Advanced in Management, Technology and Engineering Sciences ISSN NO : 2249-7455
5.3 Face Expression
Human uses distinctive facial expressions to express their sentiments or tempers. The
appearance variety brings about the spatial connection change, as well as the facial-highlight
shape change. Facial expressions can be partitioned primarily into six basic emotions, referred
to as happy, sad, disgust, surprise, fear, and anger. It is shown by different research group that
the sort of emotion expressed impacts the accuracy of any expression recognition system.
5.3 RST (rotation, scale, and translation) variation
The RST (rotation, scaling, and translation) variety is additionally caused by the variety in
image acquisition process. It brings about difficulties both in face detection and recognition,
and may require thorough searching in the detection process over all possible RST parameters.
5.4 Clutter Background
In addition to the above four variations which result in changes in facial appearances, we
likewise need to think about the impact of environments and backgrounds around individual in
images. The cluttering background influences the accuracy of face detection, and face patches
including this background additionally reduce the execution of face acknowledgment
calculations.
5.5 Occlusion
The occlusion is one of the most difficult problems in face recognition and face detection.
It means that some parts of human faces are unobserved, especially the facial features.
CONCLUSION
In this paper, we studied several factors that result in difficulties of face detection and
face recognition. Mostly the real world applications are sensitive to these factors. There are
many algorithms presented in literature to solve these problems. Further factors: illumination-
invariant, multi-view detection and recognition, and partial observation can be studied for
research work.
References
[1] Xianyang Wang , Jie Yang , XiaolongTeng , Weijun Xia , Richard Jensen, “Feature
Selection based on Rough Sets and Particle Swarm Optimization” Pattern RecognLett 28:
459- 471,2007.
[2] H. Hannah Inbarani, Ahmad TaherAzarb, G. Jothi, “Supervised hybrid feature selection
based on PSOand rough sets for medical diagnosis” Computer methods and programs in
biomedicine 1 1 3: 175–185,2014.
[3] Dinesh Kumar, Shrutika, “Harmony Search Algorithm for Feature Selection in Face
Recognition”, Computational Intelligence and Information Technology, Springer, pp 385–
390, 2011.
[4] MamtaDhanda, “Face Recognition Using Eigenvectors From Principal Component
Analysis”, International Journal of Advanced Engineering Research and Studies, Volume
1, Issue 2, pp 37-39, March 2012.
Volume 8, Issue II, FEB/2018 671 http://ijamtes.org/
International Journal of Advanced in Management, Technology and Engineering Sciences ISSN NO : 2249-7455
[5] M S Lotfabadi, A M E Moghadam, “The Comparison of Different Feature Decreasing Methods
Base on Rough Sets and Principal Component Analysis for Extraction of Valuable Features and
Data Classifying Accuracy Increasing”, International Conference on Integrated Intelligent
Computing, pp 108-113, 2010.
[6] Dinesh Kumar, Shakti Kumar, C. S. Rai, “Feature selection for face recognition: a memetic
algorithmic approach”, Journal of Zhejiang University SCIENCE A, Springer, Volume
10, Issue 8, pp 1140–1152, August 2009.
[7] Abhishek Banerjee, “Impact of Principal Component Analysis in the Application of Image
Processing”, International Journal of Advanced Research in Computer Science and Software
Engineering, Volume 2, Issue 1, January 2012.
[8] ZaidAbdiAlkareemAlyasseri, “Face Recognition using Independent Component Analysis
Algorithm”, International Journal of Computer Applications, Volume 126, Number 3, pp 34-38,
September 2015.
[9] S. Thakur, J. K. Sing, D. K. Basu, M. Nasipuri, M. Kundu, “Face Recognition using Principal
Component Analysis and RBF Neural Networks”, International Journal of Simulation: Systems,
Science and Technology, Volume 10, Number 5, pp 7-15, 2008.
[10] X Wang, J Yang, X Teng, R Jensen, “Feature Selection based on Rough Sets and Particle Swarm
Optimization”, Pattern Recognition Letters, Elsevier, Volume 28, Issue 4, pp 459-471, 1 March
2007.
[11]A Rosado-Munoz, G Camps-Valls, J Guerrero-Martinez, JV Frances-Villora, J Munoz-Mari, and
AJ Serrano-Lopez, “Enhancing Feature Extraction for VF Detection using Data Mining
Techniques” IEEE Computers in Cardiology, pp. 209-212 ,2002.
[12] MohdSaberiMohamad, SafaaiDeris, Safie Mat Yatim, Muhammad Razib Othman, “Feature
Selection Method Using Genetic Algorithm For The Classification Of Small And High
Dimension Data” First International Symposium on Information and Communications
Technologies, 2004.
[13] Ahmed Al-Ani, “Feature Subset Selection Using Ant Colony Optimization” World Academy of
Science, Engineering and Technology, 2008
[14] Chung-JuiTu, Li-Yeh Chuang, Jun-Yang Chang, and Cheng-Hong Yang, “Feature Selection
using PSO-SVM”, IAENG International Journal of Computer Science, 33:1, IJCS_33_1_18,
February 2007.
[15] Allen Y. Yang, John Wright, Yi Ma, S. Shankar Sastry, “Feature Selection in Face
Recognition: A Sparse Representation Perspective”, IEEE TRANS. PAMI, pp 1-34,
AUGUST 2007.
Volume 8, Issue II, FEB/2018 672 http://ijamtes.org/
