An Intelligent Human-Machine Interface Based

2023 International Conference on Advances in Computing, Communication and Applied Informatics (ACCAI) | 979-8-3503-1590-5/23/$31.00 ©2023 IEEE | DOI: 10.1109/ACCAI58221.2023.10200158

on Eye Tracking for Written Communication of

Patients with Locked-in Syndrome
Andril Alagusabai1, R.Sandhiya2, Murapaka Dhanalakshmi Bhavani3, Kamlesh Kukret4, T.A.Mohanaprakash5, and
A. Hajira Banu6
1
Department of Electrical and Electronics Engineering, Bannari Amman institute of technology, Sathyamangalam, Tamil Nadu
638401, India.
2
Department of Computer Science and Engineering, R M K College of Engineering and Technology, Puduvoyal, Tamil Nadu 601206,
3
Department of Electronics and Communication Engineering, Aditya College of Engineering and Technology, Surampalem, Andhra
Pradesh 533437, India.
4
Department of Electronics & Communication Engineering, Graphic Era Deemed to be University, Dehradun, Uttarakhand, India
5
Department of Computer science and Engineering, Panimalar Engineering College, Chennai, Tamil Nadu 600123, India.
6
Department of English, PSNA College of Engineering and Technology (Autonomous),Dindigul, Tamil Nadu 624622, India.

E-mail : [email protected] [email protected] [email protected]

[email protected] [email protected] [email protected]

Abstract- A deliberate loss of mobility throughout the I.INTRODUCTION

entire body, with the exception of the muscles that
control eye movements, is referred to as "locked-in Eye tracking frameworks enable the tracking of eye
syndrome" (LIS). Patients are conscious and awake
movements during a variety of tasks, including
despite this. They can communicate by blinking or
moving their eyes, and their mental faculties are feeling monitoring, reading, human movement
frequently unharmed. LIS sufferers are less recognition, viewing advertisements, visiting a
productive since their everyday routines are website, human computer interaction (HCI), driving
substantially affected and they need continual assistance frameworks, game perception, drivers'
assistance. Currently, many people use assistive exhaustion identification frameworks, and others. It
technology to make up for their limitations. In order also uses an eye-tracking system to build an "eye-
to capture eye movements, gaze courses, a client's mouse" that enhances PC usability for people with
location on the screen at any given moment, and the disabilities, and uses it as a control character to
grouping of their development, eye tracking is now frame without the user needing a mouse or other
being logically researched in Human PC Interaction.
input device. You can also enable direct interaction
This is because eye tracking technology has advanced
from an idea to the physical world. It may be argued with the work interface. I need a console. The Eye
that eye development is a crucial constant input Clan tracker consists primarily of a camera and a
technique for human-PC communication, especially high-definition infrared drive, both of which can
for those who are genuinely impaired. The suggested receive and decode signals sent from other electronic
framework aims to provide a straightforward and devices such as smartphones and computers. This
useful interactive mode using only the client's eye. The camera-based device tracks how a person's eyes
proposed framework's use stream is designed to change over time.The camera tracks even the
flawlessly correspond to human tendencies. A slightest changes in the customer's pupils by taking
magnifier module is also suggested to enable the pictures and sending them to his PC's visual
precise activity. The Innovation Acknowledgment
calculator. These calculations produce 'screen gaze
Model (Cap) measures are used to evaluate the
framework's apparent suitability. The proposed coordinates' that indicate the subject's gaze on the
framework's extreme persuasiveness in terms of screen while performing an action. This device
usability and interface design is demonstrated. works best indoors, but works well in work
environments when combined with other devices,
Keywords: Intelligent, Human-Machine, Interface, Eye camera sensors, and various lighting situations.
Tracking, Written Communication, Patients, Locked-In
Syndrome. Although both sexes, including children, can be
affected by the rare neurological condition known as

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
locked-in syndrome (LIS), adults are more likely to transmitted from a PC to a client using traditional
suffer a brain hemorrhage and a stroke. In addition, user interfaces (UIs), which give much more data
all of the body's voluntary muscles have lost their transmission capability. But, there aren't really any
movement, with the exception of those that control methods for submitting a lot of customer data in an
eye growth. Individuals with LIS are usually unable equitable fashion. HCI aims to enhance
to speak or grow their bodies, yet they are communication between users and computers by
nevertheless conscious and intellectual. They can utilizing more widespread and advantageous
communicate by blinking or moving their eyes, and communication features. Between 80 and 90 percent
their mental faculties are frequently unharmed. of the information we receive from the outside world
Patients with LIS have damage to the ventral pons, a comes from the human eye, which is one of our
portion of the brainstem that relays information to main input mechanisms. For interactive media
other regions of the brain. Due to the requirement for communication from client to PC, the eye
or dependence on pharmaceuticals as well as any developments could be considered an urgent
specific or unique aspect, this sort of condition continuous input channel. This is crucial for people
accounts for a sizeable share of a person's living with engine difficulties (like people with
expenses. Since their everyday activities are Amyotrophic Horizontal Sclerosis). How to
seriously impacted and require constant support, consistently and successfully incorporate eye
people with this illness often have lost their developments into mixed-media communication
effectiveness and personal pleasure, which may have with PC is the major goal of the study of eye
been lowered. The availability of tools and gadgets tracking strategy in client-PC communication.In
like televisions, kitchen appliances, automatic doors, general, the most logical way to incorporate eye
and assistance devices has increased thanks to advances into client PC dialogue would be to
technological innovation. Many chores around the directly replace a manual input device, like a mouse,
house can be controlled and automated. For the vast with an eye tracker. Yet, moving a real mouse with
majority of disabled individuals worldwide, skilled your hand versus utilizing your eyes to do so is
control methods to access these items can completely different. Building an eye tracking-based
significantly increase the amount of personal control framework for client PC conversation
satisfaction. requires careful consideration of the subtle but
significant differences between mouse and eye
Eye tracking innovation, which relies on eye trackers location.
to detect eye growth and position, is gaining
momentum in the fields of brain research, II.LITERATURE REVIEW
marketing, and user interfaces. Eye trackers have
been around for some time, but in the early stages of Point-of-look coordinates from an eye-look tracking
the development of the sector, their use was largely framework and electromyogram signals from the
confined to experiments in research centers, and the muscles in the face were proposed as inputs for a
use of these changes as a convenient medium of cursor control framework for PC clients by Chin et
surveillance was not. Instead, the human eye was al.It could work with a strong snap activity, but it
monitoring the idea of progress in human-PC was slower and less precise than the main control
interaction (HCI). Eye trackers he ten years ago system, which made use of eye tracking.
were prohibitively expensive and impractical for use
on real clients his computer interface. The cost is The mouse cursor is controlled by Missimer and
approximately $30,000. Well-known companies Betke's system, which also simulates the mouse's
such as Tobii, Look Point and Eye Clan Tracker left- and right-snap movements by blinking
have developed low cost eye trackers as a result of monocularly to the left or right. To determine where
recent improvements in better and more affordable the mouse cursor would be, this system used the
eye gaze interaction components. As eye tracking client's head position. The client's head's
technology becomes more accessible, new unpredictable growth would have an impact on the
applications will inevitably emerge that target the snapping ability's accuracy.
use of eye tracking in HCI.
The framework described by Lupu et al. is based on
Images, films, and other media that can quickly a device that takes and analyses pictures through a
output significant amounts of data can all be

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
camera affixed to the frame of glasses, and is III.METHODOLOGY
intended to aid in the communication of persons
with impairments.The device recognizes the patient's A product application that makes use of eye trackers
eye development, and intentional eye blinking is at a reasonable cost is the suggested eye tracking-
linked to a pictogram or word choice that reflects the based control framework. With the help of an eye
patient's needs. This framework's drawback is that tracker's "mouse cursor control" feature, the
the picture processing calculation can't reliably application can identify users' gaze. The ability to
identify acquired images of low quality and isn't adjust the mouse pointer allows users to change the
resistant to light intensity. gaze position. We will then recognize the client's
viewing location based on the location of the mouse
Using a Drove's glint and the subject's focus in a cursor. The device would generate the necessary
photograph taken with a camera, Hutchinson events if you just stared at the place for a short
calculated the eye look's bearing. If a customer looks while. Clients can select and use the appropriate
directly at the Drove, there is not much space capabilities in a similar manner.
between the glint and the student's point of focus.
On the other hand, if the client walks away from the
driver, the gap is widened. Architecture of a system
The components and relationships of the proposed
Two second request polynomial conditions were virtual mouse framework are shown in Figure 1. The
utilized by Morimoto et al. to transfer a glint-student production app is made up of the eye tracking engine
vector to a screen coordinate of a look point, and the and the mouse replication components. The eye-
coefficients of the situations were calculated by an tracking engine coordinates with the product's
adjustment approach. If there is any degree of head implementation to enable the usage of eye-tracking
development, these procedures perform better when technologies like Eye Clan Tracker, which enables
the client's head is in the right position and there is eye-control of the mouse pointer. What happens to
no identification. the data supplied by the mouse pointer is described
in the mouse updates section. As can be seen in
Using summed up relapse brain organizations, Ji and Figure 1, the various parts of this component include
Zhu suggested a brain network-based method to a primary user interface, a mouse/console emulation
arrange student boundaries to screen coordinates engine, a client activity location module, a
(GRNN). The framework makes use of student and termination (rest) module, a mouse capability
glint borders, student glint uprooting, circle fitting, module, and a console capability module.
look alignment, and mapping with student image
progressions under different quadrants, such as left,
front, and right look of the face directions. The
framework works flawlessly as long as the head is in
a good position, but accuracy suffers during free
head development.

Faisal et al. have observed that the accuracy of the

division cycle for localizing the iris structure affects
the display of a computerized iris recognition
framework. If the chosen strategy is flawed, this
could result in incorrect component extraction and
misleading recognizable proof outcomes. The
majority of developers suggest Roundabout Hough
Change (CHT) to reduce the iris limit, but this
requires extensive use of time and memory and also Figure 1:The proposed system's architecture.
an exact assessment of the limit's scope.
Workflow
The mouse cursor is used in our architecture to
control how the eyes evolve in one direction or
another. Every 50 milliseconds, the mouse

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
coordinates are measured and stored in order to track tracking until it is woken up by the user. When the
the development of the eyes. A convincing mouse coordinates deviate by 100 pixels from the
appearance (ticks) is obtained when the distance stopped coordinates, the clock begins again and an
between two close counts does not exceed 50 pixels alarm is sent to the system so that it can monitor the
in 50 consecutive counts. Customers can use it to growth of the eyes.
point and take pictures. The position of the 50th
mouse record assumes that the client should move. IV.EXPERIMENTAL RESULTS AND DISCUSSIONS
Then the current mouse coordinates were saved and
moved to the framework. An overview of the In order to evaluate the practicality and unwavering
development process of the proposed framework is quality of the framework, we carried out two tests
shown in Figure 2. First, the client's appearance is with varying degrees of complexity during the trial
distinguished by skill. Activity points are recorded stage. During the primary examination, we tested the
as her 50th mouse coordinate when a usable methodology for a series of activities, including
appearance is detected. look, duplicate, glue, etc. To assess client framework
acceptance and comprehend the rationale behind
people's acceptance or rejection of the suggested
framework, Davis' Innovation Acknowledgment
Model (Hat) was used. We then tested the client's
ability to recognize the virtual mouse while
performing a basic browsing job. Hat estimations
were used in this technique to test and address both
the initial framework and the previous framework.
The effectiveness of the suggested virtual mouse in
meeting user demands was then evaluated using the
Framework Convenience Scale (SUS). Yeah, 46
individuals took part in our analyses. They all had
eyesight that was normal or corrected to be normal,
had never heard of eye tracking, and only needed a
brief orientation to using a framework.

In order to assess how user-friendly our framework

Figure 2:The suggested system's process flow is, we gather customer feedback about it utilizing
Cap poll and SUS survey. The cap uses the activity
The toolbar pops up without warning while waiting hypothesis as a hypothetical justification for
for the client to continue selecting mouse functions. establishing the causal links between two core
In addition, gaze position models indicate whether a assumptions: saw convenience and saw usability and
consumer has seen a particular device. When you customers' mentalities, intentions, and genuine PC
select an instrument from the device bar, the reception behavior. We used the Behavioral
container for that device changes from dark to white. Intention (BI), Attitude to Use (ATU), Ease of Use
The framework also focuses on competence modules (EOU), and Ease of Use (PU) scale items from
and carries out corresponding activities. The Davis et al. accepted. And Venkatesh and Davis.
framework waits to record the position of the next The extent to which the participant acknowledged
persuasive appearance when a second persuasive that using the framework improved his presentation
appearance is needed to satisfy the device's was our definition of his PU. He called EOS the
capabilities. The coordinate data is stored and degree to which an individual perceives that using a
transferred to the functional module once the eye framework reduces their effort. ATU was
gaze recognition module gets a second convincing additionally defined as a user preference when
instantaneous coordinate. After 2.5 seconds of the interacting with the system. The possibility for
toolbar being shown, the main interface module is customers to use the framework later was BI. In
prompted to shut and store it if the framework has addition, it seems that customers will also be able to
not identified the client's preferred device view. The use it because the system will make their lives
framework then goes to sleep, disabling mouse comfortable and fulfilling. Perceived Interest (PI)

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
was thus viewed as a tool for evaluating the usability design may not be beneficial for complex search
of the framework. To estimate PI, use the elements initiatives, as evidenced by the low average PU
of Venkatesh et al. adjusted. Responses ranged from (2.60) of these designs. The first two best results of
“very different” (1) to “strongly agree” (5) on a five- mean values were obtained by PI in each case and by
point Likert scale. At the end of the meeting, ATU in the second case with values of 3.45 and
participants were invited to participate in the SUS 3.16, respectively. This indicates that the majority of
survey. This is a simple estimating tool commonly users are satisfied with the interface and
used to quantify customer opinions and frame functionality of the suggested tool. Also, the mean
preferences. It is widely used and independent of upsides of EOU and BI are both greater than three,
common engineer inventions such as equipment, indicating that members believe the proposed
programming, and location. There were 10 questions framework is user-friendly and are willing to adopt it
in the poll, with options ranging from 1 (strongly going forward.
disagree) to 5 (completely agree) (strongly agree).
The poll yields a score between one and one Table 1:TAM outcomes for the search task
hundred. It has been shown that SUS survey replies Proposed system
from members are robust, dependable, and a Constructs (n = 25) Mean SD
valuable evaluation tool. PU 2.60 2.32
EOU 3.04 2.04
A Computer with a 22-inch level screen and a 1920
x 1080 goal LCD was used in preparation for the PI 3.48 .83
testing. A program for a virtual mouse was created. ATU 3.16 2.22
A little yet incredibly accurate device called an Eye BI 3.04 2.32
Clan Tracker was used to recognize the mouse
cursor control capabilities. To guide the members in
Experiment browsing.
the examination, a summary of the exploratory
In this section, we compared the existing framework
endeavors was created.
to the preceding framework utilizing goal
confirmations as well as abstract data (from the Cap
Experiment with searching.
survey) (activity time). In order to measure user
This section serves as a test of the virtual mouse Hat
satisfaction with the system's performance while
has recommended. A brief introduction to the virtual
surfing, we also asked users to complete the SUS
mouse was given to participants before the
survey on our system. Both frameworks have been
experiment began. After that, I spent about five
applied by each participant, but in a different way.
minutes trying out the virtual mouse. Participants
Before starting the activity, participants got training
were instructed to adopt a comfortable posture and
on the framework and eye tracker alignment, similar
not move too much after exercise to increase the
to the searching trial approach. Members were
accuracy of tracking eye development. The
instructed to use one of the randomly assigned
following step was an alignment with the eye
frameworks before completing the poll for a
tracker. The group was then prepared to begin
particular framework.
resolving the errands using the suggested format.
Participants were not allowed to utilize a mouse or
Performances
any other console devices throughout the
The suggested frame hat results from the browsing
assignment; they could only use the recommended
task are shown in Table 2. Participants may feel that
application to complete the trial interaction.
an instrument is better suited for browsing than for
searching if there are more approaches to each
Participants can use left- and right-clicking, hostile
development in the browsing activity than in the
to white, scrolling, hostile to white, and virtual
search task. Similar to the search task, his two
console functionality during the interaction.
highest scores at the beginning of the browsing task
Applying the proposed structure to the search
were also given to PI (3.35) and his ATU (3.17). The
problem yields the same results as shown in Table 1.
remaining variables PU, EOU, and BI all have mean
The mean rises from 2.60 to 3.48 for each
values above 3, 4.54, 4.58, and 3.30, respectively,
development range and the standard deviation rises
indicating a positive evaluation of the proposed
from 0.83 to 2.32 for each build range. Participants
framework.
broadly agree that implementation of the proposed

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
Table 2:Outcomes of the TAM-based browsing task Table 3:comparing the lengths of time needed to
Proposed Compared activate the SCROLL function.
Constructs p
system system Participant Proposed system Compared system
(n = 20) value
Mean SD Mean SD ID (second) (second)
PU 4.54 2.03 3.84 0.89 0.024* 1. 4 6
EOU 4.58 2.05 2.42 2.07 0.256 2. 7 15
PI 3.35 0.60 3.22 0.87 0.036* 3. 3 6
ATU 3.17 0.66 2.72 0.92 0.020* 4. 4 6
BI 3.30 0.72 2.77 0.92 0.033* 5. 4 7
Average
4.3 8.7
Comparison Results Subjective Comparison time
Results.
To ascertain whether differences between the virtual System Usability Scale.
mouse frameworks had any practical importance, the Members were approached as this meeting was
Wilcox on marked rank test was performed. This coming to a close to finish the SUS poll. The results
nonparametric test was selected rather than the more of the SUS polls are shown in Table 4. The system
typical assessment of change technique due to the receives a SUS score from Bangor's analysis of
limited sample size. Table 2 illustrates the Wilcox 73.75, which is in the "Good" to "Wonderful" range,
effects on marked rank tests. According to Table 2, showing that most users think the system has a great
the suggested framework offers a significantly user interface for displaying information.
greater value at the PU.In comparison to the
examined framework, the proposed framework's Table 4:Outcomes of the SUS job of browsing
mean value is 4.54 as opposed to 3.84. It implies that System Usability Scale Mean SD
during basic operations in browsing framework, the 1. I feel like I'd like to use this 3.70 2.24
provided framework is believed to be more useful system a lot..
than the thought about framework. Moreover, 2. The system seemed overly 2.80 0.53
complicated to me.
incredibly significant contrasts were discovered for
3. The system was easy for me 3.84 0.83
variables including PI, ATU, and BI. When
to utilize.
compared to the examined framework, the proposed 4. I feel that I would need a 3.34 2.03
framework's average was 3.35, 3.17, and 3.30, as technical person's help in
opposed to the examined framework's 3.22, 2.72, order to use this system.
and 2.77, respectively. Results show that the 5. The many functions in this 4.24 0.82
suggested framework was viewed well by members system seemed to be properly
overall as an additional tool for the browsing task. interwoven, in my opinion.
6. I believed this system to be 3.20 2.03
Outcomes of an Objective Comparison far too inconsistent.
We observed the seasons of scrolling activity from 7. I anticipate that most 4.20 0.53
five members to compare the proposed framework individuals would pick up on
with the analyzed framework in order to gather using this approach fast.
8. The system was really 3.20 0.86
objective assessment data. The amount of time it
difficult for me to utilize.
takes the client to actually enable scrolling is 9. I had no trouble using the 4.00 0.75
tracked. Table 3 displays the correlation effects of system.
the startup times for both eye control frameworks 10. I had a lot of learning to do 3.24 0.82
and the Parchment capabilities. The proposed before I could start applying
framework typically opens the Parchment capability this technique.
in 4.3 seconds as opposed to the analyzed SUS score 73.75
framework's 8.7 seconds on average. It implies that
the suggested framework is more effective in Each item's standard deviation is close to or less than
browsing than other frameworks. one, indicating that the members' opinions are
comparable to one another. According to Table 4,
every approach for the unusual items is greater than
2.5, demonstrating that the majority of members are
eager to provide high and favorable scores. Which of

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
these odd questions garnered scores higher than using the framework.In contrast to the previous
three, demonstrating that the majority of respondents framework, the majority of the members greatly
thought the framework's elements were well- enjoy incorporating our virtual mouse framework
integrated and uncomplicated to use? Our into the design of the framework's interface and
framework is causing all of the methods for the even activity experiences. According to the aim proof, the
items to be lower than 2.5. This shows that the proposed framework is shown to be more effective
framework is simple to use and that the members and successful than the contrasting framework. The
can do it without a lot of specialized assistance. The participants of the SUS survey section believe that
cooperation’s generally express satisfaction with the the framework has strong integration capabilities and
suggested virtual mouse framework in their does it in an understandable way. The tasks of the
criticism. framework are also simple to understand and apply
when integrating with the Internet, sight, and sound.
V.CONCLUSION Currently, this situation is used for the standard
operating manner of acting in order to simulate a
Eye enhancement provides unbiased data on how mouse and console when interacting with a
subjects perceive the world and respond to different computer.
types of enhancement that can be used in brain
science research. Eye-tracking technology used in VI.FUTURE SCOPE
conjunction with information processing in the brain
can be based on physiological data such as: B. In the future, we'll endeavor to expand our activity
Images of the brain can be determined. Eye tracking capabilities to accommodate more client-media
is a tool that can be used to examine visual activity interaction scenarios and to adapt our framework to
and relate it to brain injury, neurological illnesses, new platforms like tablets and phones. In order to
and early stages of neurological development. Cross- provide customers with a seamless operating
referencing reading samples with various experience from turning on their computers to
socioeconomic classes of people might shed light on shutting them off, we will also foster a number of
how they gather information. Human-PC contact activity modules.
enables users to enter data onto their computers in a
more natural manner. REFERENCES
[1] Alva M., Castellino N. ,Deshpande R. ,Sonawane K. & Lopes
We provide an eye tracking-based control M. (2017). An Image Based Eye Controlled Assistive System
framework in response to the client's need to enable for Paralytic Patients. 2nd International Conference on
natural and useful PC interaction with just their eye. Communication Systems, Computing and IT Applications
(CSCITA) , 178 - 183. doi.org/10.1109/CSCITA.2017.8066549
Clients can use our framework to accomplish [2] Jabeen, Farhana, and Sarfraz Nawaz. “In-network Wireless
virtually all PC inputs without using conventional Sensor Network Query Processors: State of the Art, Challenges
and Future Directions.” Information Fusion, vol. 25, Elsevier
input hardware because it combines the capabilities BV, Sept. 2015, pp. 1–15. Crossref,
of the mouse and the console. The system gives https://doi.org/10.1016/j.inffus.2015.01.007.
typical users a new way to use a computer in [3] K. Takemura, K. Takahashi, J. Takamatsu, and T. Ogasawara,
“Estimating 3-D point-of-regard in a real environment using a
addition to enabling the disabled users to operate it head-mounted eye-tracking system,” IEEE Transactions on
in a manner similar to that of able-bodied users. Our Human-Machine Systems, vol. 44, no. 4, pp. 531–536, 2014.
Hat survey investigation revealed that the [4] L. Sun, Z. Liu, and M.-T. Sun, “Real time gaze estimation with
a consumer depth camera,” Information Sciences, vol. 320, pp.
participants thought our eye development framework 346–360, 2015.
was not difficult to understand. Members express [5] M. A. Eid, N. Giakoumidis, and A. El Saddik, “A novel
eyegaze-controlled wheelchair system for navigating unknown
interest in looking at and reading content utilizing environments: case study with a person with ALS,” IEEE
the suggested eye control framework in the interim. Access, vol. 4, pp. 558–573, 2016.
They are keen to learn more about the outcomes of [6] M. Nehete, M. Lokhande, and K. Ahire, “Design an eye
tracking mouse,” International Journal of Advanced Research in
our study on how to interact with a computer Computer and Communication Engineering, vol. 2, no. 2, 2013.
utilizing eye tracking. [7] N. Scott, C. Green, and S. Fairley, “Investigation of the use of
eye tracking to examine tourism advertising effectiveness,”
Current Issues in Tourism, vol. 19, no. 7, pp. 634–642, 2016.
The suggested structure enhances surfing [8] O. Ferhat and F. Vilarino, “Low cost eye tracking: the current
effectiveness and user experience in the browsing panorama,” Computational Intelligence and Neuroscience, vol.
2016, Article ID 8680541, pp. 1–14, 2016.
test. Customers can easily interact with media by [9] Oken, B. S., Orhan, U., Roark, B., Erdogmus, D., Fowler, A.,

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.
 Mooney, A., & Fried-Oken, M. B. (2014). Brain–computer [14] Shneiderman. (2017). designing the user interface: strategies for
interface with language model–electroencephalography fusion effective human-computer interaction. Pearson Education India.
for locked-in syndrome. . Neurorehabilitation and neural repair, [15] Y.-M. Cheung and Q. Peng, “Eye gaze tracking with a web
28(4), 387-3. doi.org/10.1177/1545968313516867 camera in a desktop environment,” IEEE Transactions on
[10] Q. Sun, J. Xia, N. Nadarajah, T. Falkmer, J. Foster, and H. Lee, HumanMachine Systems, vol. 45, no. 4, pp. 419–430, 2015.
“Assessing drivers’ visual-motor coordination using eye [16] S B G TilakBabu and ChSrinivasa Rao, "An optimized
tracking, GNSS and GIS: a spatial turn in driving psychology,” technique for copy–move forgery localization using statistical
Journal of Spatial Science, vol. 61, no. 2, pp. 299–316, 2016. features", ICT Express, Volume 8, Issue 2, Pages 244-249,
[11] R. G. Lupu, F. Ungureanu, and V. Siriteanu, “Eye tracking 2022.
mouse for human computer interaction,” in Proceedings of the [17] S B G TilakBabu and ChSrinivasa Rao, "Efficient detection of
4th IEEE International Conference on E-Health and copy-move forgery using polar complex exponential transform
Bioengineering (EHB ’13), Iasi, Romania, November 2013. and gradient direction pattern" , Multimed Tools Appl (2022).
[12] Rahate, Anil, et al. “Multimodal Co-learning: Challenges, https://doi.org/10.1007/s11042-022-12311-6.
Applications With Datasets, Recent Advances and Future
Directions.” Information Fusion, vol. 81, Elsevier BV, May
2022, pp. 203–39. Crossref,
https://doi.org/10.1016/j.inffus.2021.12.003.
[13] S. M. A. Meghna, K. L. Kachan, and A. Baviskar, “Head
tracking virtual mouse system based on ad boost face detection
algorithm,” International Journal on Recent and Innovation
Trends in Computing and Communication, vol. 4, no. 4, pp.
921– 923, 2016

Authorized licensed use limited to: Amrita School of Engineering. Downloaded on February 11,2025 at 06:58:41 UTC from IEEE Xplore. Restrictions apply.