Re-orchestration of Remote Teaching Environment in eLearning

Uma Gopalakrishnan, Ramkumar N, P. Venkat Rangan and Balaji Hariharan

Amrita Center for Wireless Networks and Applications, Amrita School of Engineering,

Amrita Vishwa Vidyapeetham University, Kollam, India

Keywords:

Knowledge Source, Learning Experience, eLearning, Gesture Correlation.

Abstract:

eLearning has become indispensable in making lectures from experts accessible to a wide student audience.

Current eLearning solutions offer many ways to re-create the teaching environment in a remote location. They

focus on recreating the elements that make up the teaching side. In this paper we identify the components

of a teaching environment and the factors that constitute a conducive learning environment for students. We

identify the common methods of teaching side reconstruction and conduct a study to evaluate the effect of

each on the learning outcome on these factors. We analyse the network and computational performance of

these methods and arrive at a most optimal method to use in a eLearning session.

1 INTRODUCTION

ELearning has become essential in making quality ed-

ucation reach the masses. Students at remote places

can attend live lectures from experts who are at a

physically distant place. Moreover, over the years,

the process of knowledge delivery by a teacher has

transformed and got enriched. Teaching has no longer

remained the teacher just lecturing and using plain

black board. It now very often includes the teacher us-

ing digital content such as presentation, internet based

material etc and has impacted eLearning technolo-

gies in a multitude of ways (Lau et al., 2014). This

had made the task of recreating a realistic immersive

experience of knowledge delivery at a remote place

complex for live lectures. How effective the eLearn-

ing systems are largely depends on how best each sys-

tem is able to recreate the teachers side, including its

multitude of components, in remote places.

We deﬁne the components of the teaching side

which constitute the content delivered by the teacher

as Knowledge Sources. From our observation of

many live teaching sessions, the following Knowl-

edge sources were identiﬁed to constitute a regular

classroom lecture:

• Teaching material - Teaching material used by the

professor which is presented on the white board

or presentation screen.

• Teachers audio - Teachers speech containing the

explanation of the topic

• Teachers video Teachers video gives a feeling

of physical presence of the teacher and also is a

source of information which consists of the fol-

lowing components: a) Coordinated gestures -

Gestures used by the teacher such as hand move-

ment, cursor movement etc which relates the con-

tent spoken by the teacher to the teaching material

displayed. b) Independent gestures Facial expres-

sions and other gestures used by the teacher spon-

taneously when he/she speaks to explain a topic.

All these components need to be present to create

a complete learning environment. There are multiple

ways in which these knowledge sources can be rep-

resented digitally for transport to the remote location.

This creates multiple media types for the knowledge

transport. How these media types can be combined

and represented in the remote location gives multiple

possibilities for knowledge representation in the re-

mote site.

1.1 Problem Description

The eLearning and video conferencing solutions like

A-VIEW, Saba, Skype, Google hangouts etc enable

some of the possibilities for teachers side representa-

tion. In a simplistic scenario, some systems display

video of the teachers part of a class containing the

teacher, whiteboard etc. The video from the teach-

ers physical location is compressed and streamed real-

time to the remote location. Here, we often see that

when the teacher is explaining a concept using a pre-

sentation or writing something on the board, the con-

Gopalakrishnan, U., N, R., Rangan, P. and Hariharan, B.

Re-orchestration of Remote Teaching Environment in eLearning.

In Proceedings of the 18th International Conference on Enterprise Information Systems (ICEIS 2016) - Volume 2, pages 223-229

ISBN: 978-989-758-187-8

223

tent become indiscernible over the video.

To counter this some systems allow uploading the

teaching material beforehand for remote students to

access or the screencast of the digital content used by

the teacher is streamed to the remote location along

with teachers audio. In either case when the remote

students listen to the teacher lecturing live, in many

instances they dont understand which part of the dig-

ital content to view in order to follow the teacher.

When both teachers video and the screencast is

streamed, these two need to be in synchronization.

These components are represented as separate distinct

entities in the remote classroom which makes it difﬁ-

cult for the remote students to correlate between the

two. When the teacher is pointing at some content on

the board while explaining the topic, it becomes dis-

tracting for the remote students as they have to ﬁrst

understand the correlation between the teachers video

and the teaching content before they can focus on the

topic being discussed.

Each of these techniques gives a different experi-

ence of the learning environment to the remote stu-

dents and is effective to different extent. Each tech-

nique also differs in cost and performance associated

to it in terms of the network bandwidth utilized, la-

tency and complexity of software and the hardware

conﬁguration involved. Current research lacks in giv-

ing a scientiﬁc method to evaluate the learning expe-

rience given by these techniques.

1.2 Our Contribution

In this paper, we analyze various factors of a real

classrooms knowledge delivery process that needs to

be recreated and preserved in an eLearning session

for an effective learning experience. We develop a

systematic scientiﬁc way to evaluate and quantify the

effect of each method of recreating the teaching side

on the learning experience of the students based on

these factors. We start from a most basic method

of representation of knowledge sources, evaluate its

effects, and proceed towards more complex methods

by adding missing factors step by step and ﬁnally ar-

rive at an optimal method of representing the teaching

components which results in the best learning experi-

enced by the remote students. As we go from sim-

ple to complex methods there is also an increase in

the cost associated with recreating the multiple media

types in terms of network bandwidth and latency. We

measure the changes in bandwidth utilization and la-

tency in the system and examine the changes in the

complexity of the system. We present the results on

both enhancement in the learning experience and the

cost incurred due to the increased complexity. As we

present these results, we discuss whether the cost and

complexity involved in an optimal system is afford-

able and whether the gain in learning experience jus-

tiﬁes the additional complexity and cost of the system.

This work is part of a bigger project NKN (Na-

tional Knowledge Network) gateway to global class-

room, which aims at making the eLearning experi-

ence immersive for the students and teachers. The

NKN (NKN, 2012) is a multi-gigabit network across

India which provides high speed network backbone

for connecting Educational institutions across India.

We take advantage of this super high speed network

for taking eLearning experience to the next level.

The rest of the paper is organised as follows. Sec-

tion 2 gives an overview of related work. Section

3 describes the experiment design and experimental

setup. Section 4 analyses the results of the experiment

and Section 5 gives the conclusion.

2 RELATED WORK

ELearning technologies have gained wide popular-

ity because of its capability to give quality education

to masses. There has been a lot of study to gauge

the effectiveness of eLearning systems. Tanzila Saba,

in her paper (Saba, 2012) examines the interconnec-

tion between eLearning systems, self-efﬁcacy and the

learning outcome for online courses. A case study

performed on the blackboard system, a web-based

eLearning system (Liaw, 2008) analyses the factors

affecting the learning efﬁciency and found that stu-

dents self-efﬁcacy, use of multimedia for instruction

and interactive modules in the system had a positive

outcome on the learning. Studies on blended learning,

i.e, blending online media into a course show that it

enhances the learning outcomes of the students (Wai

and Seng, 2015). MOOC platforms like edX, Udac-

ity, Coursera etc have become hugely popular as a re-

sult (Round, 2013). Most of the studies so far have

been based on the efﬁciency of such online courses

and evaluating the effectiveness of eLearning systems

which act as content provider through various multi-

media forms and networking capabilities.

However, a large community of students and

teachers use eLearning in its synchronous form where

the teaching session is live and there is direct face

to face interaction between students and the teacher.

This gives the students and teachers a better sense

of participation in the classroom session (Hrastinski,

2008).

Video conferencing tools like Skype (Sivula,

2011), Google hangouts (Sean Gallagher, 2013) and

Cisco telepresence (Cisco, 2012) offer a ready way

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

224

for live face to face interaction between students and

teacher. Apart from sharing video content for face to

face communication, these tools also offer facility to

share the screen and thus share any teaching material

which is being discussed. There are dedicated eLearn-

ing platforms as well like A-VIEW (Lab, 2008), Saba

(Saba, 2015) and other similar systems which provide

features speciﬁc to eLearning in addition to enabling

face to face interaction. They enable personal interac-

tion through video and chat, content sharing, screen

sharing, students participation with polls and white-

board activity.

Current research lacks in measuring the effective-

ness of such synchronous live eLearning methods.

Most of study had been based on measuring the ef-

fect of self paced methods on the learning outcomes

and the best possible ways to design them. This pa-

per aims to evaluate the learning outcomes and per-

formance of some live synchronous techniques of

eLearning.

3 EXPERIMENT DESIGN

The experiments are designed by identifying the most

basic way to reconstruct the components of teaching

in the remote site, analyze its effect and the missing

factors and then upgrading the experiments by design-

ing progressively more complex methods using differ-

ent combinations and representation of the knowledge

sources.

3.1 Representing Knowledge Sources

The knowledge sources of a regular classroom are

teaching material, Teachers audio, Teachers video

consisting of coordinated gestures and independent

gestures. When these sources need to be translated

into media types for transport to the remote location,

we come across multiple forms of media representa-

tions. Some representations include multiple knowl-

edge sources whereas others represent single source.

The media type representations are as follows:

• Video of the Environment (V

Env

) - Video contain-

ing the teacher and the teaching material.

• Teachers Video (V

) Teachers zoomed in video,

without the teaching material.

• Screencast (V

) Video of the activities on the

presentation screen containing the teaching mate-

rial.

• Pointer Video (V

SC−P

) - Video of the activities

on the presentation screen containing the teach-

ing material along with a pointer added to it for

indicating the area teacher is focusing on.

• Teachers Audio (A

) Audio of teachers speech.

3.2 Reconstruction Techniques

Representation of the Teachers Side

View

The different combinations of the knowledge sources,

their representations, relative positioning and size at

the remote classroom give rise to different possibili-

ties of recreating the teachers side at the remote site.

We start from a most basic representation in our anal-

ysis and enrich the representations using multiple me-

dia types as follows. In this process, we chose to se-

lect those methods which are representative of a real

classroom situation and avoided representation such

as communication using text chat which are not natu-

ral in a real classroom.

• Environment Video - <V

Env

, A

>. A video

stream of the teachers area that includes in its

view the teacher, teaching material used in a class-

room is displayed in the remote classroom. This

forms the most basic representation of the compo-

nents. Figure 1 depicts the scenario.

Figure 1: Environment Video representation.

• Screencast - <V

, A

>. Screencast of the digital

content used as teaching material by the teacher is

streamed and displayed in the remote classroom

along with the teachers audio. Figure 2 shows the

view seen by the remote students.

Figure 2: Screencast at the remote classroom.

• Screencast and Teachers video - same screen

, V

, A

>. Screencast of the digital con-

tent used as teaching material by the teacher is

Re-orchestration of Remote Teaching Environment in eLearning

225

streamed and displayed in the remote classroom

along with the teachers audio and teachers video

on the same screen as the screencast. The teachers

video consists of the teacher zoomed in without

the full view of the teaching material as shown in

Figure 3.

Figure 3: Screencast and Teachers video.

• Screencast and Environment video - separate

screen <V

, V

Env

, A

>. Screencast of the digital

content used as teaching material by the teacher is

streamed and displayed in the remote classroom.

Along with this the teachers video including the

presentation screen content in the background is

displayed on a separate screen as shown in Figure

Figure 4: Environment video and Screencast.

• Screencast, Teachers video and a pointer - same

screen <V

SC−P

, V

, A

>. Screencast of the

digital content used as teaching material by the

teacher is streamed and displayed in the remote

classroom along with the teachers audio and

teachers zoomed in video on the same screen as

the screencast. Here, whenever the teacher refers

to the presentation screen a pointer in the screen-

cast indicates the content area being focused as

depicted in Figure 5.

3.3 Experimental Setup

The set up consists of a local classroom, where the

teacher is physically present and a remote classroom,

where the remote students are attending the lecture

live.

Figure 5: Screencast, pointer and teacher’s video.

3.3.1 Hardware Components

• Each classroom included a Logitech C920 HD pro

camera connected to a PC (conﬁguration) to cap-

ture and stream the video of the participants.

• The local classroom consisted of a Sony 52 inch

TV for the teacher to view the remote students and

the TVs inbuilt speakers were used for the remote

students audio

• The participants were given Shure directional

mics to capture their audio and stream though a

PC.

• The local classroom consisted of a smart board

(Smart Technologies) which the teacher could

use as a white board as well as presentation

screen. The smart board content were captured

and streamed.

• The remote classroom consisted of two Sony 52

inch TVs to display the teacher and other media

whenever required.

3.3.2 Software Components

• Gstreamer was used as the streaming solution for

video and audio. This was chosen for its speed.

• Smart technologies software for smart board func-

tionalities and touch feature.

• VLC media player was used for streaming the

screencast

4 EVALUATION AND ANALYSIS

Based on the knowledge sources representation and

the effect each source has on the learning outcome,

we arrived at the following factors that contribute to-

wards an effective learning experience. Along with

the effect of knowledge sources, we also evaluate how

each method inﬂuences the interactivity of the stu-

dents with the teacher:

• Presence: Teachers physical presence plays a vi-

tal role in engaging the students in the learning

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

226

process. In addition, it is also a direct source of

knowledge when the teacher uses some gestures

to convey information.

• Clarity: This factor refers to the visual clarity of

the information content on the white board or pre-

sentation screen.

• Interaction Alignment: Refers to the teachers in-

teraction direction, i.e, who the teacher is address-

ing to while lecturing and whether the students

feel the teacher is addressing them.

• Ease of Interaction: Ease with which the students

and teacher are able to interact with each other.

This factor captures whether the students were

able to catch the attention of the teacher and talk

to him as easily as if the teacher was physically

present.

• Gesture Correlation: Understanding the correla-

tion between professors gaze, gestures and posi-

tion with the teaching material focused on.

• Learning Experience: This factor summarizes the

learning experience based on how well the stu-

dents were able to focus on the content delivered

in the class.

4.1 Learning Experience Evaluation

The reconstruction techniques were evaluated for

these factors by conducting a study where each of

these techniques were employed and experimented

with one by one in a remote classroom and a survey

was taken to measure these factors. One topic was

taught for each of the reconstruction technique. The

topics were chosen such that they are similar in their

complexity level and involved using material which

had lots of images, equations which are necessary for

explanation. The teacher had to often point at the con-

tent and the students needed to view the content to

understand the topic better. The teaching sessions in

each experimentalso involvedinteraction between the

remote students and the teacher.

Same teacher taught all the sessions to maintain

uniformity in the lecture delivery to avoid any distor-

tions in the result. The remote students were asked to

rate each factor on a scale ranging from 0 to 10. Here

0 denoted total absence of the factor and 10 denoted

the factor being present as it would in a real classroom

to its full extent. Table 1 shows the results.

4.2 Performance Evaluation

The video streams from the local and remote class-

rooms were sent at a SD resolution of 640x480 at

25 fps. The screencast was streamed at a HD resolu-

tion of 1920x1080 (display resolution) at 20 fps. The

video streams are encoded using X264 encoder, and

the audio stream using mp3 encoder. Corresponding

decoding is done at the receiver. Latency in the sys-

tem is introduced because of the encoding/decoding

process and the delay in transmission through the

network. The classrooms were connected through 1

Gbps LAN network.

The performancein terms of bandwidth utilization

and latency in the system are presented in Table 2.

We observe that the increase in latency is due to

HD streaming of the screencast and its associated en-

coding and decoding. However, the latency was ob-

served to be tolerable by the participants. The im-

provement achieved in the learning experience be-

cause of the usage of screencast overshadowed the

slight inconvenience caused by the latency. The band-

width utilization also did not hamper the performance

of the systems. The experiments were conducted over

a local area network setup. Over distant places, we

propose to utilize the ultra-high speed NKN. This

would eliminate the problems arising out of band-

width requirement to a great extent.

In addition to measuring these factors, subjective

analysis of the techniques was done and the discus-

sion on our ﬁnding is as follows:

• Environment Video - <V

Env

, A

> This tech-

nique was close to the natural real classroom in

terms of the relative positioning and sizes of the

knowledge sources and to a certain degree all the

knowledge sources were recreated. This tech-

nique also fared best in terms of the feeling of

presence of the teacher and the ability to inter-

act with the teacher. The main drawback was

brought out when we measured the clarity of the

digital teaching material. When any digital con-

tent is used by the teacher and the remote students

view the content through its video, its clarity is

degraded to such an extent that in many cases stu-

dents were not able to follow the topic. This loss

in clarity renders this technique unusable in situ-

ations where the teacher uses any digital teaching

material.

• Screencast - <V

, A

> To compensate for the

loss of clarity in the teaching material, screencast

of the digital content was streamed to the remote

classroom. A simplistic reconstruction was done

where the students view the teaching material and

listen to the teachers speech. The teachers phys-

ical presence was lost completely and it affected

the ease of interaction as well. The students could

not relate the audio to its source and found it difﬁ-

cult to interact with the teacher. Also, the students

Re-orchestration of Remote Teaching Environment in eLearning

227

Table 1: Rating of the factors affecting the learning outcome.

Techniques Presence Clarity Interaction

Alignment

Ease of Inter-

action

Gesture cor-

relation

Learning

Experience

Env

, A

> 7.3 5.33 8.5 9.5 8.33 5

, A

> 0.66 9.66 1.5 5 0 3.5

, V

, A

> 5.33 9.66 6 6.5 6.33 7

, V

Env

, A

> 6.33 9.33 6.5 7 7 6.5

SC−P

, V

, A

> 5.33 9.66 6 6.5 8 7.5

Table 2: Network performance of the reconstruction tech-

niques.

Techniques Bandwidth

Utilization

Latency

Env

, A

> 5 Mbps + 77

kbps

250 ms

, A

> 5 Mbps + 77

kbps

1 s

, V

, A

> 7.5 Mbps + 77

kbps

1 s

, V

Env

, A

> 7.5 Mbps + 77

kbps

1 s

SC−P

, V

, A

> 7.5 Mbps +

77kbps

1 s

were required to put strained effortsin the absence

of teachers video and gestures for correlating the

teachers speech to the teaching material. Some-

times the teachers audio alone was insufﬁcient to

understand where to look at the content on the

white board or presentation screen. The teacher

had to be stopped for clarifying and made the pro-

cess of knowledge delivery interrupted with such

distracting breaks. In spite of the presence of

these difﬁculties, this method was better usable

than the ﬁrst technique where there was not much

scope of understanding the digital content.

• Screencast and Teachers Video - Same Screen

, V

, A

> The teachers missing presence

is included to some extent by adding the teach-

ers video in a corner on the same screen as the

screencast. Since the size of the video displayed

is very small, the teacher is not seen as clearly as

in technique 1 which reduces the feeling of phys-

ical presence in comparison. Though the gestures

of the professor are seen in the video, the correla-

tion between the gestures and the teaching content

cannot be made out easily. This again makes fo-

cusing on the content difﬁcult.

• Screencast and Environment Video - Separate

Screen <V

, V

Env

, A

> To see the teacher

clearly including the gestures and its related con-

tent, the teachers side video as in technique 1 is

displayed along with the screencast on separate

screens. This is intended to assist the student in

correlating the gestures and the digital content and

enhance the teachers presence. However, the sur-

vey results show that the students found it very

distracting and difﬁcult to follow two screens for

the same content. It was very inconvenientto keep

switching the gaze between the two screens. Most

of the students surveys preferred technique 3 in

comparison.

• Screencast, Teachers Video and a Pointer -

Same Screen <V

SC−P

, A

> To improve upon

the students perception of the correlation between

teachers gestures and the teaching material, a

pointer is used in the screencast. Rest of the repre-

sentation remains same as in technique 3. When-

ever, the teacher points to any content by touch-

ing the smart board or uses mouse movement to

indicate some region of content, the pointer im-

age appears on the screencast which aids in the

students understanding what content to focus on.

The teaching delivery becomes easy to follow for

the students. The survey results show this tech-

nique as the most effective. Even though the

teachers presence is felt to a limited extent, this

technique results in the best learning experience

by the students.

5 CONCLUSIONS

Re-construction of the teachers side in remote class-

rooms is the most fundamental aspect in eLearning.

We have identiﬁed the most used techniques for the

reconstruction and evaluated each technique for its

performance and the learning experience it provides.

We started by experimenting with a most basic form

of reconstruction where a video of the teachers side

is presented to the remote students along with the

teachers audio. This required minimal hardware and

software setup and had good performance. Though

this was the best way to create the teachers pres-

ence, it lacked some vital factors for giving a good

learning experience to the students. In order to bet-

ter the students learning experience, we experimented

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

228

with other techniques to present the teachers side and

arrived at a most optimal way for doing so. The

technique where the screencast, teachers video and a

pointer to correlate teachers gestures and the teaching

material were shown in the same screen was evalu-

ated as providing the best learning experience with

a rating of 7.5 against 10. This technique fared best

when we assessed the clarity and the correlation be-

tween teachers gestures and the content presented on

screen. There was an increase in the cost with respect

to the bandwidth utilized and latency which was how-

ever tolerable to a large extent. Our future work will

aim at bettering the learning experience of the stu-

dents by enhancing the naturalness of the recreated

environment.

ACKNOWLEDGEMENTS

We would like to express our gratitude to Mata

Amritanandamayi Devi (Amma), Chancellor, Amrita

Vishwa Vidyapeetham, for being a constant source of

inspiration and guidance. We also thank the partic-

ipants who made it possible to conduct the experi-

ments and carry out the research.

REFERENCES

Cisco (2012). Cisco telepresence.

http://www.cisco.com/c/en/us/products/collaboration-

endpoints/immersive-telePresence/index.html.

Accessed: 2016-02-16.

Hrastinski, S. (2008). The potential of synchronous com-

munication to enhance participation in online discus-

sions: A case study of two e-learning courses. Infor-

mation & Management, 45(7):499–506.

Lab, A. E.-L. R. (2008). Video conferencing tool, a-view

- the versatile e-learning tool. http://aview.in/. Ac-

cessed: 2016-02-16.

Lau, R. W., Yen, N. Y., Li, F., and Wah, B. (2014). Re-

cent development in multimedia e-learning technolo-

gies. World Wide Web, 17(2):189–198.

Liaw, S.-S. (2008). Investigating students perceived sat-

isfaction, behavioral intention, and effectiveness of

e-learning: A case study of the blackboard system.

Computers & Education, 51(2):864–873.

NKN (2012). National knowledge network. http://nkn.in/.

Accessed: 2016-02-16.

Round, C. (2013). The best mooc provider: A review

of coursera, udacity and edx. Web. http://www.

skilledup.com/blog/the-best-mooc-provider-a-review-

of-coursera-udacity-and-edx.

Saba (2015). Virtual classrooms.

http://www.saba.com/us/solutions/virtual-

classrooms/. Accessed: 2016-02-16.

Saba, T. (2012). Implications of e-learning systems and

self-efﬁciency on students outcomes: a model ap-

proach. Human-Centric Computing and Information

Sciences, 2(1):1–11.

Sean Gallagher, C. J. (2013). Google beefs up

hangouts into text, photo, video chat power-

house. http://arstechnica.com/information-technolo

gy/2013/05/google-beefs-up-hangouts-into-text-

photo-video-chat-powerhouse/. Accessed: 2015-12-

30.

Sivula, M. (2011). Using skype as an academic tool:

Lessons learned.

Wai, C. C. and Seng, E. L. K. (2015). Measuring the ef-

fectiveness of blended learning environment: A case

study in malaysia. Education and Information Tech-

nologies, 20(3):429–443.

Re-orchestration of Remote Teaching Environment in eLearning

229