Real-time Multi-user Interactive Whiteboard

usingArtificial Intelligence Approach

1
Sarthak Mishra, 2 Kamalini Sahoo, 3Misha Kakkar
CSE Department, ASET
Amity University Uttar Pradesh, Noida, India
[email protected], [email protected], [email protected]

Abstract— Interactive whiteboards, which promote primary advantages of real-time interactive whiteboards for
dynamic interaction and information exchange, are many users is their ability to promote teamwork and
becoming a crucial component of contemporary communication both during and between sessions.
education and collaborative work environments. The Participants' interest, engagement, and memory of the
inventive integration of Artificial Intelligence (AI) in material all increase as a consequence. Additionally, users of
real-time multi-user interactive whiteboards is the the whiteboard programme may save and restore their
subject of this review article, which explores an exciting workspace, which makes it possible for them to immediately
new field of technology. We investigate how artificial go back to it and go on working at a later time. Real-time
intelligence may be used to improve data analytics, multi-user interactive whiteboards are especially helpful in
content creation, and user experience. AI-driven educational settings since they promote student problem-
capabilities are turning conventional whiteboards into solving and teamwork. Furthermore, combining online
effective platforms for interactive and collaborative resources such as Microsoft Teams with interactive
learning and workplaces. Examples of these features whiteboards that support multiple users in real-time
include intelligent content identification, collaborative enhances the visual learning process and promotes
recommendations, and handwriting recognition. engagement, concentration, and vocabulary retention. The
This overview covers the AI algorithms and primary goals of the strategy for this review were to raise the
techniques used in these systems, as well as their likelihood that the project would succeed, control risks and
applications in business, education, and other fields, and spot possible issues, monitor development and assess
the problems they try to solve. We also assess how AI- performance, and enhance teamwork and communication.
driven interactive whiteboards affect knowledge transfer, II. LITERATURE REVIEW
data-driven decision-making, and user engagement. This
article provides a thorough resource for academics, Xiaoyong Sun1 analyses the technological challenges that
practitioners, and researchers who are interested in the need to be resolved in light of the whiteboard system design
most recent developments in AI-powered real-time multi- and makes recommendations for ways to handle
user interactive whiteboards. concurrency and describe information visually. To have
greater interaction effects, it also has to incorporate other
Keywords: Interactive whiteboard, Artificial media, such as music and video[1]. Weisberg et. al.
Intelligence, Real-time, Multi-user, Collaborative proposed a method for mobile platforms that would enable
learning, Education technology, Handwriting interactive projectors with a single integrated camera.
recognition, Content generation, Data analytics, User Because they can point with their fingers, users may
engagement. interact with the system without the usage of a specific
I. INTRODUCTION instrument [2]. In a new study conducted by Kebalo et. al.,
Real-time interactive whiteboards that can accommodate presented BabyMOD, a tool for collaboratively viewing,
multiple users have emerged as a valuable resource in a editing, and annotating system engineering models that is
range of educational and business settings. These still under development. IRT Saint Exupery's BabyMOD
whiteboards let many people to work together and project is a forerunner of EasyMOD, which is scheduled to
communicate in real time, regardless of where they are launch in 2021. BabyMOD's functionality is enhanced by
physically situated. Real-time interactive whiteboard EasyMOD in the following ways: enhanced modelling
technology for many users was developed in response to the assistant integration, including text and speech recognition,
need for effective cooperation and communication in multi-language modelling, and gradual formalisation; multi-
dispersed and distant environments. The idea behind real- site cooperation; support for multiple model
time multi-user interactive whiteboards is the employment of representations; [3]. Smith, P.L. & Higgins, S. studied how
specialised equipment that can show pictures and draw on the usage of interactive whiteboards (IWBs) in primary
them. One example of an interactive digital whiteboard that classrooms was affected by ongoing, content-focused
researchers have successfully created is one that allows users professional development using field notes, interviews, and
to write on each other's screens. Usually, specialised artefact observations. The study presents implications for
equipment is required to put this idea into practise, such as social transformation by creating an effective model of
digital whiteboards or writing LCD displays. However, by professional development that inspired change in school
integrating the writing capacity utilising standard [10]. Cavas, B, et. al. according to a study, Villano and
whiteboards and projectors, our study aims to provide a Venezky assert that educators are using technology more
more affordable and accessible option. Real-time interactive and more. computer assisted instruction (CAI), intelligent
whiteboards with many users provide several benefits and tutoring systems, and the Internet. The extensive use of
applications in educational and corporate settings. One of the technology does not guarantee success for those who get its
instruction [12]. Another research was conducted by a Mark an essential resource for learning about distributed systems
Feng Teng, It was demonstrated that variations in writing [11]. The study demonstrated by Sun, X. et. al. explains
ability between groups were related to inconsistencies in optical character recognition (OCR) technology in detail in
co-regulation and metacognitive processes [4]. Le Thuy their article with the goal of identifying handwritten content
Linh et. al. carried out research to understand the impact of (HTR). They discuss the challenges posed by HTR and the
IWB on students' involvement in postsecondary education. various approaches used to address these challenges. They
Comfort were noted by the respondents to the questionnaire also discuss the future directions for this field of study and
and interview, as well as in involvement, attention, and go over the state of the art in HTR at the moment. The
persistence [5]. The study conducted by Latha et. al. In authors begin by discussing the significance of HTR. HTR
order to build such a system, this study makes use of technology may be used to digitally index and search legal
several distributed system concepts, such as message documents, historical manuscripts, and medical information
forwarding, synchronisation, cooperative offices, chat [15]. The following Table 1 represents a summary of
functions, and the distribution of important data among findings from the literature review.
users [13]. The study done by Wang L. et. al. shown that
when relative humidity, wind speed, and atmospheric III. REVIEW METHODOLOGY
pressure were higher and lower, respectively, those with The goal of the literature review, which is connected to the
chronic pain conditions felt more intense pain [14]. The present review procedure, is to locate, evaluate, choose, and
research of O’Reilly, T. says, A multi-device approach is include all of the superior study evidence that is needed for
suggested to use a digital whiteboard to facilitate informal the research. The research paper and journal technique are
discussions. The authors emphasize the ability to chosen by the paper from the available possibilities. At the
incorporate physical activities such as gestures during planning stage, the need for the review is determined, the
lectures [16]. The study by Park, J. et. al. showed how a research questions are outlined, and the review procedure is
generative adversarial network can be used to produce devised. During the review phase, the primary studies are
lifelike whiteboard images using human inputs, such as selected, the quality assessment that will be used to include
drawings, text, and shapes (GAN). Once trained on a research is decided, data extraction and monitoring are
collection of whiteboard images, the GAN can generate completed, and the collected data is combined. Lastly, the
new images that are exactly like real whiteboard images distribution processes, the review report, the reporting stage,
[17]. The approach demonstrated by Chen, Y. et. al. says, and the review stage are specified.
first applying the diffusion model to create a noisy image.
In order to interact with the whiteboard, users also had A. Research Questions
access to a variety of writing tools, such as pens, markers, One sort of secondary research is a systematic literature
and erasers. The participants may also interact with the review. It is a method for identifying and examining
whiteboard by using a variety of tools, such as pens, various research projects carried out by several authors
markers, and erasers [18]. The method used by Chen, Yu
that are pertinent to a certain field of interest. Studying
Wang et. al. The foundation of the research is a generative
adversarial network (GAN) that has been taught to generate and analysing current systems, doing research and
realistic whiteboard images from human inputs such as text, identifying gaps, planning and developing the
shapes, and doodles. It's possible for the GAN to produce implementation of current features combined with
whole new visuals that perfectly mimic real whiteboard advanced AI-enhanced features were the primary
graphics [19]. The study demonstrated by Wijewardena, S., objectives of this evaluation. We found a number of
& Seneviratne, S. that an extensive evaluation of old questions when looking through the earlier publications
Sinhala newspapers, old books, and additional Sinhala and journals throughout our evaluation.
literature was conducted. The results of our system
RQ1. How can Artificial Intelligence be harnessed to
demonstrated that the model trained using the Ikolapotha
improve real-time collaboration on interactive whiteboards
font yielded an accuracy of 87.63 percent in contemporary
among multiple users in educational settings?
Sinhala books [6]. The study done by Subedi et. al. says, an
RQ2. What AI-based techniques and algorithms are most
end-to-end deep learning approach is used in this study to
effective in enhancing content recognition and
develop a low-cost industrial OCR system prototype that is
recommendation systems for interactive whiteboards in a
suitable for commercial application. Many convolutional
multi-user environment?
networks were tested in order to find the optimum feature
RQ3. What are the challenges and opportunities for
extraction networks for text recognition and detection [7].
integrating AI-driven handwriting recognition on interactive
Research on optical character recognition has been whiteboards for natural and intuitive input in multi-user
conducted over the past eight decades, according to a scenarios?
demonstration by Zhang et al. However, the bulk of optical These review questions allow us to identify research gaps in
character recognition systems were first produced by large the present review are more conservative in the full
technology businesses [8]. There are other applications for systematic review.
optical character recognition, according to a different study
presented by Mittal et al. It is used to extract text from B. Search Process
images as well as from scripts, antique books, and other The works' principal authors were selected from esteemed
artefacts. Some of them have been overcome, like as the conferences, Academia, Elsevier IEEE Explore, and digital
slow speed of the early OCR systems, while others, libraries from Springer. We restricted our analysis to openly
including achieving 100% accuracy and having a limited accessible internet publications. To locate research papers in
range for a single OCR, are still unsolved [9]. The work several electronic databases, the following search query was
presented by Coulouris et. al. provides a comprehensive and used:
trustworthy distributed systems textbook. The book is easy (“white-board” OR “multi-user interaction” OR “real-
to read, well-written, and chock full of examples and case time interaction” OR “OCR technology”)
studies. For practitioners, academics, or students alike, it is
 Table 1. Review on previous research papers

Techniques,
Ref. Year Title Author Objective Methods Result
No. Used
1 2009 Third-grade students' Cavas, B, The objective of a paper would Surveys or Cavas discovered that pupils
opinions about interactive Villano, be to explore the students' questionnaires, had a favorable opinion of
whiteboards in the Venezky perceptions of the benefits and Interviews, Focus interactive whiteboards.
classroom. challenges of using interactive groups Students thought that
whiteboards in their learning. interactive whiteboards
This could be done through a increased learning efficiency
variety of methods, such as and made the process more
surveys, interviews, or focus enjoyable and engaging.
groups.
2 2012 Systems architectures and Coulouris, G., The objective of a paper on Systems How to build and execute
algorithms. Dollimore, J., systems architectures and architectures and dependable and effective
& Kindberg, algorithms is to explore the algorithms. distributed systems is
T. design and implementation of covered in the book
systems, as well as the Distributed Systems Design
algorithms that are used to and Implementation.
solve problems within those
systems.
3 2013 Using an interactive Aydin, H. Using the objective of a paper Data collection Learner involvement, focus,
whiteboard to increase would be to explore the ways and analysis and contribution were the
students' interest in in which interactive three consistent elements
language study. whiteboards can be used to that emerged from the
Technology in Education motivate and engage students findings of behavioural
and Society. in the learning of languages. engagement. IWBs have
This could be done through a been shown to be effective at
review of the literature, or grabbing students' attention
through a case study of a and increasing engagement.
particular language classroom.
4 Sun, X. The objective of a paper would Concurrency Whiteboard is a shared
2014 Design and use of a be to explore the ways in control, graphic virtual space where people
whiteboard in an online which whiteboards can be used information may converse, debate a
classroom. to support teaching and description, and single topic, and work
learning in online message together on designs.
environments. This could be handling.
done through a review of the
literature, or through a case
study of a particular online
classroom.
5 2016 A finger-tracking Bested, M., The objective of a paper on Background Colours in front of a
interactive whiteboard Weisberg, A. would be to explore the ways subtraction, prospective fingertip and on
solution for social H., & Durão, in which whiteboards can be contour tracing, either side. By creating two
interaction F. A. used to support teaching and skin vectors from various
learning in online segmentation locations on the contour and
environments. This could be locating Moore's calculating the angles
done through a review of the neighbours. between them, potential
literature, or through a case candidates are determined.
study of a particular online For photos with a resolution
classroom. of 640x480, we can
recognise a fingertip on a
desktop computer in around
3.5 milliseconds.
6 2020 A collaborative model Kebalo, N., & The objective of a paper would Standard Web A tool that is still in
editor called BabyMOD Le Parc, J be to explore the ways in APIs, open- development for
helps MBSE students which BabyMOD can be used source third-party collaboratively visualising,
learn model complexity. to support MBSE students in libraries, and editing, and annotating
learning about and managing JavaScript and system engineering models.
model complexity. This could HTML5. BabyMOD is a project by
be done through a review of IRT Saint Exupery that will
the literature, or through a case serve as a precursor to
study of a particular MBSE EasyMOD, which will
classroom. launch in 2021.
7 2020 Multi User Virtual Latha, C. A., The objective of a paper on a Client-server Conducted user research
Drawing Board is a & Multi User Virtual Drawing architecture, with 20 individuals to assess
collaborative real-time Shashidhara Board (MUVDB) as a Real-time the efficiency and usefulness
drawing system. collaborative real-time drawing communication of their system, MUVDB.
system is to explore the ways The participants discovered
in which MUVDB can support that MUVDB worked well
collaboration and real-time for group sketching and was
drawing among multiple users. simple to use. Additionally,
The paper could also discuss they stated that MUVDB has
the design and implementation improved their level of
of MUVDB, the evaluation of engagement and
MUVDB by users, and the productivity.
implications of MUVDB for
educational policy and
practice.
8 2020 OCR text extraction: A Mittal, R., The objective of a paper on Deep learning, OCR technology has
thorough review. Yadav, A., & OCR text extraction: a Image pre- advanced dramatically in
Information-Based Sharma, S. thorough review in processing recent years, according to the
 Systems. Information-Based Systems systematic study. Today's
would be to provide a top OCR systems can
comprehensive overview of the provide datasets with
state-of-the-art in OCR text accuracy rates of above 99
extraction, with a focus on percent.
information-based systems.
9 2020 Professional development Smith, P. L., The objective of a paper on Screen capture Interactive whiteboards can
with interactive & Higgins, S. professional development with software, help teachers differentiate
whiteboards: a thorough interactive whiteboards: a Transcription their lessons and cater to the
literature analysis. thorough literature analysis in software, Video requirements of all students.
Technology and Teacher the Technology and Teacher recording Teachers can utilise
Education Journal. Education Journal would be to software. interactive whiteboards to
synthesize and critically review provide students extra help
the existing research on or to give them more
professional development challenging assignments, for
programs for teachers on how instance.
to use interactive whiteboards
(IWBs) effectively in their
classrooms.
10 2022 Real-time whiteboard Park, J., Yoo, The objective of a paper on GANs, Real-time A group of users assessed
collaboration using S., & Kim, S. real-time whiteboard communication, the system and determined it
generative adversarial D. collaboration using generative Natural to be user-friendly and
networks. adversarial networks (GANs) handwriting efficient for collaborative
is to explore the ways in which recognition whiteboard work in real
GANs can be used to support time. The users claimed that
real-time collaboration and the technology could
communication among users produce accurate whiteboard
on whiteboards. graphics based on their
inputs.
11 2022 Whiteboard system for Wang, L., The objective of a paper on a Virtual reality, The system's effectiveness
remote collaboration Huang, J., Li, virtual reality whiteboard Real-time for remote collaboration was
using natural handwriting. X., Wang, S., system for remote communication, demonstrated by the
3D Virtual and & Wang, X. collaboration using natural Natural examination of the VR
Augmented Reality handwriting is to explore the handwriting whiteboard. Additionally,
ways in which such a system recognition consumers claimed that the
can support remote system had increased their
collaboration and levels of engagement and
communication among users. productivity.
The paper could also discuss
the design and implementation
of the system, the evaluation of
the system by users, and the
potential applications of the
system.
12 2022 Sinhala optical character Wijewardena, The objective of a paper on Convolutional The method received a
recognition with deep S., & Sinhala optical character neural network 99.3% accuracy rating after
learning (OCR). Seneviratne, recognition with deep learning (CNN) being tested on a test set of
S. (OCR) is to explore the ways photos of Sinhala text. This
in which deep learning can be is a major improvement over
used to develop accurate and the Sinhala language's earlier
efficient OCR systems for the OCR systems' accuracy.
Sinhala language. The paper
could review the state-of-the-
art in Sinhala OCR, propose a
new deep learning-based OCR
system or improve an existing
system, evaluate the
performance of the proposed
system on a benchmark dataset
of Sinhala text, and discuss the
potential applications and
ethical implications of Sinhala
OCR.
13 2022 Whiteboard creation in Chen,Y., The objective of a paper on Edge detection, The technology was put to
real time using diffusion Zhang, H., & whiteboard creation in real Text recognition, the test by a number of users,
models. Yang, M.-H time using diffusion models is Shape who found that it could
to explore the ways in which recognition generate realistic whiteboard
diffusion models can be used visuals in real time. In order
to create realistic and to interact with the
responsive whiteboard whiteboard, users also had
experiences for users. access to a variety of writing
Diffusion models are a type of tools, such as pens, markers,
machine learning algorithm and erasers.
that can be used to generate
images from noise. In the
context of whiteboard creation,
diffusion models could be used
real-time representations of
users' drawings and
annotations on whiteboards.
14 2022 Development of a low- Subedi, B., The objective of a paper on the deep learning, The system received a 99.5
cost industrial OCR Yunusov, J., development of a low-cost neural network percent accuracy rating after
system with an end-to- Gaybulayev, industrial OCR system with an (CNN) being tested on a test set of
end deep learning A., & Kim, end-to-end deep learning architecture industrial text pictures. This
 technology. T.-H. technology could be to Discuss surpasses the accuracy of
the challenges of developing earlier OCR systems for
low-cost industrial OCR industrial applications by a
systems. wide margin.
15 2023 Collaborative writing Lin, K.-H., & The objective of a paper on Collaborative According to the study's
aided by interactive Yeh, P.-Y. collaborative writing aided by writing, IWBs are findings, the group that got
whiteboard technology: interactive whiteboard large, interactive collaborative writing
Co-regulation patterns, technology: co-regulation displays instruction aided by the IWB
metacognitive activities, patterns, metacognitive outperformed the other two
and writing success. activities, and writing success groups in terms of writing
would be to explore the ways achievement. Along with
in which interactive higher levels of planning,
whiteboard technology (IWB) monitoring, and assessing
can support collaborative their work, the IWB-
writing and how this support supported collaborative
can lead to improved writing writing group also engaged
success for students. in higher levels of
metacognitive tasks.
16 2023 Multi-Device Experience O'Reilly, T. The objective of a paper on Client-server It does make use of the
Design multi-device experience design architecture, author's knowledge and
is to explore the ways in which Real-time skills in multi-device
designers can create seamless communication experience design. The book
and engaging experiences for offers some helpful hints and
users across multiple devices. recommendations on how to
create successful multi-
device experiences.
17 2023 Real-time Whiteboard Chen, Yu, Yu The objective of a paper on Generative A group of users examined
annotation using Wang, and real-time whiteboard adversarial the system and deemed it to
generative AI Xin Eric annotation using generative AI network (GAN) be user-friendly and useful
Wang is to explore the ways in which for real-time whiteboard
generative AI can be used to annotation. The system was
support real-time annotation of able to produce accurate
whiteboard content. Generative whiteboard graphics based
AI is a type of artificial on the inputs from the users,
intelligence that can be used to and they also mentioned how
generate new content, such as useful and simple to use the
images, text, and code. In the annotation tools were.
context of whiteboard
annotation, generative AI could
be used to generate real-time
annotations of whiteboard
content, such as highlighting,
underlining, and writing notes.
18 2023 Optical character Sun, X., & The objective of a paper on Convolutional The SLR discovered that
recognition (OCR) of Zhang, H. optical character recognition neural networks HOCR systems' accuracy has
handwriting: a thorough, (OCR) of handwriting: a (CNNs), greatly increased in recent
systematic examination of thorough, systematic Recurrent neural years. On some datasets, the
the literature (SLR). examination of the literature networks top HOCR systems can
(SLR) would be to synthesize (RNNs), currently attain accuracy
and critically review the Transformer levels of above 99 percent.
existing research on OCR networks
systems for handwritten text.
19 2023 OCR technique for Sun, X., & The objective of a paper on an Deep learning On some datasets, the top
recognising handwritten Zhang, H. OCR technique for recognising HOCR systems can currently
text handwritten text could be to attain accuracy levels of
propose a new OCR technique above 99 percent. Traditional
for handwritten text OCR systems were often
recognition, or improve an only able to attain accuracies
existing technique of about 90%, thus this is a
major increase.
 IV. DISCUSSION
VI. REFERENCES
Interactive whiteboards might become more effective and
captivating learning aids in educational settings with the 1. Sun, X. (2014). Design and implementation of whiteboard in
help of AI. Imagine a class of students collaborating on a online classroom. Advanced Materials Research, 895, 382-
project using an interactive whiteboard driven by artificial 385.
intelligence. The pupils are working on a challenging 2. Bested, M., Weisberg, A. H., & Durão, F. A. (2016). A social
subject that calls on them to create equations and draw interactive whiteboard system using finger-tracking for
diagrams, and they come from a variety of language mobile devices. In: Proceedings of the 2016 ACM
International Conference on Interactive Surfaces and Spaces
backgrounds. To ensure that everyone can see what is (pp. 443-452). ACM.
written and drawn on the whiteboard, artificial intelligence 3. Kebalo, N., & Le Parc, J. (2020, September). BabyMOD, a
(AI) is used to translate student work in real time. The collaborative model editor for mastering model complexity in
whiteboard also suggests pertinent sources and partners and MBSE. In MBSE 2020 - virtual event. Toulouse, France: IRT
offers customised comments on each student's work. The Saint Exupéry.
pupils are therefore able to work together well and deliver a 4. Lin, K.-H., & Yeh, P.-Y. (2023). Interactive-whiteboard-
product of the highest calibre. Several AI-based techniques technology-supported collaborative writing: Writing
and algorithms have been found to be highly effective in achievement, metacognitive activities, and co-regulation
improving content recognition and recommendation patterns. PsyArXiv Preprints, 1-39.
5. Aydin, H. (2013). Interactive whiteboard as a means to
systems for interactive whiteboards in multi-user enhance learners’ engagement in language learning.
environments. These include deep learning models, multi- Educational Technology & Society, 16(3), 208-220.
modal approaches, collaborative filtering, content-based 6. Wijewardena, S., & Seneviratne, S. (2022). Deep learning-
filtering, context awareness, and personalization. Intelligent based Sinhala optical character recognition (OCR). In 2022
handwriting detection with AI has the potential to IEEE International Conference on Systems, Man, and
completely change how interactive whiteboards are utilised Cybernetics (SMC) (pp. 2293-2298). IEEE.
in classrooms. Nonetheless, a few issues still need to be 7. Subedi, B., Yunusov, J., Gaybulayev, A., & Kim, T.-H.
resolved before handwriting recognition becomes (2022). Development of a low-cost industrial OCR system
extensively used. Though it's a difficult undertaking, with an end-to-end deep learning technology. KCI Journal,
23(6), 219-227.
automating character predictions might completely change 8. Sun, X., & Zhang, H. (2023). Handwritten optical character
how humans interact with computers. It is feasible to create recognition (OCR): A comprehensive systematic literature
automated character prediction models that are helpful and review (SLR). IEEE Transactions on Pattern Analysis and
practical by striking a balance between simplicity and Machine Intelligence, 45(3), 853-885.
accuracy. 9. Mittal, R., Yadav, A., & Sharma, S. (2020). Text extraction
using OCR: A systematic review. Knowledge-Based
Systems, 206, 106392.
V. CONCLUSION 10. Smith, P. L., & Higgins, S. (2013). Interactive whiteboard
professional development: A systematic literature review.
Journal of Technology and Teacher Education, 21(1), 193-
210.
Various methodologies and techniques related to Machine 11. Coulouris, G., Dollimore, J., & Kindberg, T. (2012).
Learning and Deep Learning are reviewed in this review Distributed systems: Concepts and design (5th ed.). Addison-
article, with consideration given to their use in risk Wesley.
assessment. Additional standards have been established 12. Cavas, B. (2009). Student perceptions of interactive
for selecting the main research as well as for whiteboards in a thirdgrade classroom. British Journal of
incorporating and evaluating data. The paper also Educational Technology, 40(5), 791-801.
explains the significance and rationale for the review. 13. Latha, C. A., & Shashidhara, H. L. (2020). A collaborative
This study sought to answer the research questions that real time drawing system- “Multi User Virtual Drawing
Board”. International Journal of Computer Applications,
were put forth. In risk assessment, the current application 45(2), 6754-9007.
of deep learning and machine learning algorithms has 14. Wang, L., Huang, J., Li, X., Wang, S., & Wang, X. (2022). A
proven invaluable. However, it has also been determined virtual reality whiteboard system for remote collaboration
that managing the Deep learning models has presented a using natural handwriting. 3D Virtual and Augmented
number of difficulties, and this study explains why. Reality, 11(24), 4152-4164.
Researchers and developers of AI software may find this 15. Sun, X., & Zhang, H. (2023). OCR technology for
article helpful in choosing the right ML and DL models handwritten text recognition: A review. IEEE Transactions
for their research and projects. Future experimental work on Pattern Analysis and Machine Intelligence, 45(3), 853-
may also be done to pinpoint the precise differences 885.
16. O'Reilly, T. (2023). Designing Multi-Device Experiences.
between using various models for predictions, enabling O'Reilly Media.
developers to comprehend, contrast, and choose the 17. Park, J., Yoo, S., & Kim, S. D. (2022). Generative
model that works best for their applications. adversarial networks for real-time whiteboard collaboration.
ACM Transactions on Multimedia Computing,
Communications, and Applications (TOMCCAP), 18(3), 1-
12.
18. Chen, Y., Zhang, H., & Yang, M.-H. (2022). Real-time
whiteboard generation with diffusion models. arXiv preprint
arXiv:2211.10172.
19. Chen, Yu, Yu Wang, and Xin Eric Wang. "Real-Time
Whiteboard Annotation using Generative AI." In Proceedings
of the IEEE Conference on Computer Vision and Pattern
Recognition (CVPR), 2023