Analyzing Machine Learning Techniques for Facial Recognition Project

This project delves into the application of machine learning techniques for facial recognition. It begins with an introduction to the topic, outlining the project's aim, objectives, and research questions, which focus on analyzing the usage of facial recognition, identifying issues with existing techniques, and evaluating the role and methods of machine learning in improving facial recognition systems. The literature review covers facial recognition techniques, including adaptive regional blend matching and generalized matching, and discusses the use of machine learning algorithms like support vector machines and neural networks. The research methodology section details the qualitative research approach, interpretivism philosophy, exploratory research design, primary data collection through surveys, and non-probability sampling. Data analysis will be conducted using thematic analysis. The project also includes a plan outlining the timeline for completion, covering introduction, literature review, data collection, and conclusion phases, along with a list of references.

PROJECT

TABLE OF CONTENTS
INTRODUCTION...........................................................................................................................1
Description of the project topic...................................................................................................1
Title..............................................................................................................................................1
Aim of the project........................................................................................................................1
Objectives of the project..............................................................................................................1
Research questions.......................................................................................................................1
LITERATURE REVIEW................................................................................................................2
Theme 1 Facial recognition technique.........................................................................................2
Theme 2s Use of machine learning in facial recognition............................................................2
RESEARCH METHODOLOGY....................................................................................................3
Research type...............................................................................................................................3
Research philosophy....................................................................................................................3
Research design...........................................................................................................................4
Data collection.............................................................................................................................4
Sampling......................................................................................................................................4
Data Analysis...............................................................................................................................5
PLAN...............................................................................................................................................5
REREFENCES................................................................................................................................7

INTRODUCTION
Description of the project topic
Automatic human fa e detection is one of the most common biometric knowledge which
is being used in most of the system, devices etc. This biometric technique has gathered lot of
attention recently. This technique is used for identification of faces whether in still images, in
videos or for unlocking a device by recognizing faces. Adaptation of this technology has
increased drastically and there are many applications that have adopted this technique and are
being used for face detection (Khan and et. al., 2019). Not only this many mobile companies
have also implanted this feature of face detection and unlocking within their devices. This
technique is used for uniquely identifying and verifying each and every person. However, there
are various kinds of issues that are associated with current approaches that are being used for
facial detection and in order to overcome these issues machine learning is being used for facial
recognition. This project will majorly focus upon ways in which machine learning can be used
for facial recognition.
Title
Ways in which machine learning can be used for facial recognition
Aim of the project
The main aim of this project is “To analyse ways in which machine learning can be used
for facial recognition”.
Objectives of the project
Main objectives of this project are as follows:
 To analyse usage of facial recognition technique
 To identify issues associated with existing facial recognition techniques
 To analyse use of machine learning in facial recognition
 To evaluate machine learning techniques that can be used for facial recognition
Research questions
Main research questions of this project are as follows:
 What are the ways in which face recognition technique is being used?
 What are the main issues associated with existing facial recognition techniques?
 How machine learning can be used in facial recognition?
 What are the main machine learning approaches that can be used for facial recognition?
1

LITERATURE REVIEW
Theme 1 Facial recognition technique
According to the view of Finizola and et. al., (2019) Facial recognition is a kind of
biometric technique which is majorly used for identification and verification of a person by
comparing and analysing their facial contours. This technique can be used for identification of
different kinds of people whether in real time, in pictures or in videos. This technique is being
used by different kinds of organization for different purposes. Not only this, various kinds of
systems or devices are being developed within in- built facial recognition features. For example,
it is being used within mobile devices for unlocking, it being pre-installed within security
cameras for identification of faced, it is also being used within law enforcement software’s used
by police for identification people. Other than this it is also used within mobile applications for
identification of faced from images or videos. Usage of this technique is increasing drastically
but there are many issues associated with this technology. Many other different languages or
methods are being worked upon so that these issues can be resolved and because of this, various
kinds of research are taking place so that all kinds of issues associated with this technique can be
resolved and features of this technology can be enhanced in a much better manner. Finizola and
et. al., (2019) further explains that, there are different kinds of facial recognition techniques that
are being used currently such as adaptive regional blend matching method, generalized matching
face detection method and many more. Most of these techniques are based upon different nodal
points upon face of a human. These measurement values are being used against variable points
on face of a human. However, there are many issues that are being associated with these existing
techniques such as many techniques used for facial detection fail at recognizing colour of people,
some of the fail in differentiating between twin faces of two different peoples. In order to
overcome these issues, advance technology such as machine learning can be used.
Theme 2s Use of machine learning in facial recognition
As per the view of Koneru and et. al., (2018) machine learning is an application of
artificial intelligence that helps the system in providing an ability to learn things automatically
and bring improvement within its explicit programming from past experience. Machine learning
is being used in various kinds of applications, games etc. This technology can be used for facial
recognition software’s or applications as well. Machine learning algorithms takes large datasets
as input and learn from this data. Most of the machine learning algorithms goes though the data
2

and tries to identify a pattern from the data. Machine learning can help in resolving various kinds
of issues or problems associated with facial detection in existing technologies. With the help of
this technology there is no need of development of new neural network, machine learning
algorithm can learn from existing database. If an image is passed, then that image is compared to
existing data stored and if the image is similar or close enough with existing image only then
output is displayed. Koneru and et. al., (2018) further explains that, machine learning usage is
increasing rapidly in artificial intelligence world. It is not only effective and being used for
games, but it is also an effective technique and is being used for face detection or recognition as
well. It is an effective facial recognition technique because it automatically learn and recognize
different and complex patterns and analyse them with existing data and based on the analysis
results it takes decisions automatically. Due to this increasing usage and implementation of facial
detection an recognition technique different kinds of machine learning techniques are being used
for example: support vector machines, HMM model, , neural network and many more. Each of
these machine learning technique or algorithm has its own way of detecting faces and is used in
different kinds of languages.
RESEARCH METHODOLOGY
Research type
There are two main types of research that helps in carrying out a research in a proper and
systematic manner. First type of research is qualitative type of research and second is
quantitative type of research. Qualitative research is based upon non- numeric and unquantifiable
data i.e. data collected cannot be analysed using mathematical techniques (Kumar, 2019).
Whereas, Quantitative type of research is based upon resolving a problem using numbers i.e. all
the numerical data is analysing using mathematical methods. For this research qualitative type
of research will be used as it will make is much easier for the researcher to collect and analyse
data collected for the research in order to answer research questions in a proper and appropriate
manner.
Research philosophy
Research philosophy helps in understanding ways in which data should be gathered,
analysed and used. There are two main types of research philosophies that can be used within a
research that are: interpretivism and positivism. Interpretivism research philosophy is used when
human interest is required to be integrated into a study. This philosophy is mostly used in
3

qualitative type of research. Whereas positivism philosophy is used for adhering factual
knowledge which is gained though observation. It is mostly used in quantitative type of research.
For this research project interpretivism philosophy will be used as it is a qualitative type of
research.
Research design
There are three types of research design: exploratory, descriptive and casual. Exploratory
is used for exploring specific areas of research, descriptive research design is used for describe
specific elements of research area (Mohajan, 2018). Whereas, casual is used for to study cause-
and-effect relationships. For this research exploratory research design will be used as it will help
in exploring main aim and objectives of the research in such a manner that research question can
be answered in a much elaborative manner.
Data collection
Data collection is one of the most important part of research project which helps in
collection of information that can be analysed in order to answer the research questions in a
much better manner. there are two types of data collection method: primary and secondary data
collection method. Primary data is collected by the researcher themselves with the help of
methods such as interview, survey, observation etc. Whereas secondary data is a kind of data
that has already been published and can be obtained from books, Journals, articles and from
many other ways. For this research, research will be using primary data collection method in
which data will be collected with the help of a survey.
Sampling
Sampling is another important part of a project which is used for selecting members from
a population in order to be included in the study. It helps in making a research type and size
manageable, helps in saving overall time of the project (Ørngreen and Levinsen, 2017). There
are two methods in which though which sample size of a research project can be selected. First is
probability method in which every members of a population have a known chance of
participating within the study. Whereas, non- probability sampling method is used when only
selected members have a chance to participate within a study i.e. not every member has a chance
of participating within a study. For this research project non- probability sampling method will
be chosen for selecting sample population 20 individuals using machine learning and work in
Mobile companies.
4

Data Analysis
Data Analysis is a systematic process which is used for analysis of data collected with the
help of statistical or logical techniques or methods for answering the research question or
reaching to the conclusion in a proper manner. there are two types of data collection methods:
thematic and statistical data collection method. Thematic data collection method us used for
segregating data in common themes. Whereas, statistical data analysis method is used for
analysis of numerical data using statistical operations. For this research project, researcher will
be using thematic data analysis method as it is one of the most appropriate type of analysis
method that can be used for analysis of qualitative data in a proper manner.
PLAN
Activity 1st
week
2nd
week
3rd
week
4th
week
5th
week
6th
week
7th
week
8th
week
9th
week
10th
wee
k
Introduction
Formation
of aim and
objectives
Research
proposal
Literature
review
Data
collection
Evaluation
of project
outcomes
Conclusion
and
recommenda
tions
5

Submission
6

REREFENCES
Books and Journals
Finizola, J.S., and et. al., 2019, July. Comparative study between deep face, autoencoder and
traditional machine learning techniques aiming at biometric facial recognition. In 2019
International Joint Conference on Neural Networks (IJCNN) (pp. 1-8). IEEE.
Khan, S., and et. al., 2019, March. Facial recognition using convolutional neural networks and
implementation on smart glasses. In 2019 International Conference on Information
Science and Communication Technology (ICISCT) (pp. 1-6). IEEE.
Koneru, N., and et. al., 2018. Surveillance Camera with Facial Detection and Recognition using
machine learning. International Journal of Pure and Applied Mathematics. 118(20).
pp.3961-3967.
Kumar, R., 2019. Research methodology: A step-by-step guide for beginners. Sage Publications
Limited.
Mohajan, H.K., 2018. Qualitative research methodology in social sciences and related
subjects. Journal of Economic Development, Environment and People. 7(1). pp.23-48.
Ørngreen, R. and Levinsen, K., 2017. Workshops as a Research Methodology. Electronic
Journal of E-learning. 15(1). pp.70-81.
7