SlideShare a Scribd company logo
1 of 24
AUTOMATIC CHEATING DETECTION
IN EXAM HALL
Mr.PRAVEEN
ASST PROFESSOR)
ROOPIKHA.S-200701201
ROSHINI.S-200701202
RITHIKA.S-200701198
Department of Computer Science and Engineering
CS19811 – Project Phase-II
Problem Statement and Motivation
The Aim is to develop an automatic cheating detection system using machine
learning techniques to identify and prevent cheating behavior during exams
conducted in an exam hall environment this system can detect various forms of
cheating such as copying from materials, communicating with others using electronic
devices or other suspicious behavior while minimizing false positive and negatives
cheating during exams determines the integrity of the education system and
diminishes the value of academic achievement by using automatic cheating detection
system provides more efficient and accurate solution by leveraging advance
technologies like computer vision and data analysis this system ensure fairness and
maintain the exam protocol.
Phase-II First Review Department of Computer Science and
Engineering
2
Objectives
Implementing an automatic cheating detection system in exam halls using machine
learning has the following objectives:
1. Enhancing Exam Integrity
2. Minimizing False Positives
3. Ensuring Scalability
4. Reducing Human Errors
5. Achieving Cost and Resource Efficiency
These objectives aim to provide a comprehensive and reliable solution to maintain the
credibility of exams and uphold the principles of academic honesty.
Phase-II First Review Department of Computer Science and
Engineering
3
Abstract
An innovative cheating detection system for exam halls using machine learning is
being introduced. By utilizing computer vision and real-time monitoring, the system
can identify various cheating behaviors, thus ensuring exam integrity. This model's
training minimizes errors and false alerts, optimizing accuracy.
Automated alerts help administrators take prompt interventions to uphold fairness and
credibility. The system also adheres to privacy regulations, safeguarding student data.
This innovative system enhances cheating prevention and maintains the essence of
academic honesty.
Phase-II First Review Department of Computer Science and
Engineering
4
System Architecture
Phase-II First Review Department of Computer Science and
Engineering
5
System Architecture
Existing System
The existing methods for cheating detection in exam halls often rely on manual
monitoring by invigilators, which can be time-consuming and prone to human
errors.these methods include: visual inspection, surveillance cameras ,proctoring
software .these, methods have limitations such as limited coverage, potential privacy
concerns and inabilities to detect subtle cheating methods.
Proposed System
The proposed system for automatic cheating detection in exam halls will use
YOLOv3 with ShuffleNets to integrate cutting-edge technologies that enhance
accuracy, efficiency, and ethical considerations in monitoring exam environments.
Phase-II First Review Department of Computer Science and
Engineering
6
System Architecture
The proposed system is made up of several components that work together to
maintain the integrity of an exam hall. These components include the integration of
the YOLOv3 object detection algorithm, which identifies cheating behaviors or
objects such as smartphones, notes, or collaboration between students with high
accuracy. Additionally, the system uses the ShuffleNets architecture to optimize
computational efficiency, enabling real-time detection while reducing computational
resources. A network of strategically-placed cameras provides comprehensive
coverage for video input to the system. To accurately distinguish between normal
activities and potential cheating incidents, the system is trained to classify specific
cheating behaviors. It continuously monitors the exam hall through live video
analysis and immediately flags any suspicious activity through an alert mechanism
developed to notify proctors or administrators in real-time.
Phase-II First Review Department of Computer Science and
Engineering
7
System architecture
System Operation:
The system is designed to adapt to varying lighting conditions, camera angles, or
exam hall layouts for consistent and reliable performance. The system operates
ethically by respecting individual privacy and focusing solely on relevant cheating
behaviors without unnecessary intrusion. A user-friendly interface is developed for
proctors or administrators, allowing them to monitor the system, review flagged
instances, and take appropriate actions effectively. In operation, the proposed system
continuously analyzes live video feeds from multiple cameras using YOLOv3 with
ShuffleNets, enabling real-time detection of cheating behaviors. Upon identifying
suspicious activities, the system promptly alerts administrators via the user interface,
where they can review flagged instances and take appropriate actions to maintain the
integrity of the examination process.
Phase-II First Review Department of Computer Science and
Engineering
8
List of Modules
1. Clearly define the objectives and goals of your project:
What exactly do you want to achieve with automatic cheating detection using
YOLOv5
2. Identify the Stakeholders:
Identify all the key stakeholders, including end-users, project sponsors, and anyone
else who will be impacted by or involved in the project.
3. Gather Requirements:
Collect and document the specific requirements for the cheating detection system.
These may include data sources, performance metrics, and user interface
requirements.
Phase-II First Review Department of Computer Science and
Engineering
9
List Of Modules
4. Data Collection and Annotation:
Determine what data you will need for training and testing your YOLOv5 model.
This includes images or videos of test-taking environments. Annotate the data to
identify cheating activities.
5. Model Selection:
Decide on the use of YOLOv5 as the object detection framework. Assess its
suitability for your specific project requirements.
6. Hardware and Software Requirements:
Identify the hardware and software infrastructure needed to run YOLOv5 efficiently.
Consider GPU/CPU resources, storage, and software dependencies.
Phase-II First Review Department of Computer Science and
Engineering
10
List Of Modules
7. Team Formation:
Assemble a team with the required skills, such as machine learning engineers, data
scientists, and software developers.
8. Project Timeline:
Create a detailed project timeline that outlines milestones, deadlines, and
deliverables.
10. Data Preprocessing:
Clean, preprocess, and augment the data as necessary for training and testing the
YOLOv5 model. Ensure data privacy and security considerations are met.
Phase-II First Review Department of Computer Science and
Engineering
11
Functional Description for each modules with
DFD and Activity Diagram
Phase-II First Review Department of Computer Science and
Engineering
12
Implementation/Results of Module
Project Description for the First Module:
Objective:
The primary objective of the first module is to develop an automatic cheating
detection system using YOLOv3 object detection with ShuffleNets architecture,
specifically designed for monitoring and preventing cheating behaviors in an exam
hall or similar testing environments.
Key Features of Functionalities:
1. Real-time Object Detection: Our system uses YOLOv3 (You Only Look Once) to
provide accurate object detection in real-time.
Phase-II First Review Department of Computer Science and
Engineering
13
Implementation/Results of Module
2. Integration of ShuffleNets: Our system utilizes ShuffleNets architecture to
optimize and streamline the detection process. This enables efficient use of
computational resources while maintaining accuracy.
3. Detection of Cheating Behaviors: We have developed algorithms to detect cheating
behaviors such as looking at unauthorized materials, communication between
candidates, or suspicious actions.
4. Integration with Exam Monitoring System: Our cheating detection system is
integrated with an exam monitoring setup to trigger alerts or notifications when
suspicious activities are detected.
Phase-II First Review Department of Computer Science and
Engineering
14
Implementation/Results of Module
5. Data Logging and Reporting: Our system captures and logs instances of suspicious
behavior for further review and analysis by authorized personnel.
This module is designed to cater to the needs of educational institutions, examination
centers, or any organization that conducts exams where maintaining the integrity of
the testing process is of utmost importance. The primary goal of this module is to
assist invigilators or administrators in managing and monitoring exam halls to ensure
that examinations are conducted fairly and by the established guidelines.
Technologies and Tools:
Phase-II First Review Department of Computer Science and
Engineering
15
Implementation/Results of Module
We have implemented the YOLOv3 deep learning model to detect objects in real
time. To optimize computational efficiency without compromising accuracy, we have
employed the ShuffleNets architecture. Python is the primary programming language
we use for implementation. Additionally, we have leveraged popular deep-learning
libraries such as TensorFlow and PyTorch for model training and inference.
The estimated timeline for the completion of the first module is as follows: Research
and Design: 1 month, Model Development and Integration: 2 months, Testing and
Refinement: 1.5 months, Integration with Exam Monitoring System: 1 month.
Phase-II First Review Department of Computer Science and
Engineering
16
Implementation/Results of Module
The initial module's outcomes aim for robust cheating detection in exam halls using
YOLOv3 with ShuffleNets. Expected results include high accuracy in identifying
cheating behaviors (potentially 90%), efficient real-time performance (averaging 0.1
seconds per frame), and minimized false positives (around 5%) and false negatives
(about 8%). Successful integration with exam monitoring systems, triggering alerts in
95% of detected cheating instances, is crucial. Preliminary performance metrics such
as precision (91%), recall (88%), and F1-score (89%) will be indicators. These
outcomes serve as the foundation for subsequent improvements, guiding further
iterations. Any shortcomings will prompt retraining with more diverse datasets,
adjustments in model parameters, or architectural modifications to optimize the
system's performance. User feedback and real-world testing will drive refinements in
subsequent modules for a more reliable and comprehensive cheating detection system
.Phase-II First Review Department of Computer Science and
Engineering
17
Pending Modules & Conclusion
The proposed model was implemented using Python with Tensor Flow on a core i7
system with Nvidia 1080 ti, and trained with a learning rate of 0.0002, batch size of
1, for 100 epochs, and momenta of 0.091. The training started with a loss of 4.02323,
which gradually reduced to 0.0023. The model was trained on 75% of the total
dataset, with 15% used for validation and 10% for testing. The proposed model
achieved an accuracy of 88.03% on our dataset, while the YOLOv3 model with
transfer learning scored 82.06% accuracy. The proposed model performed better than
the typical YOLOv3, both in terms of time and accuracy. Below Table shows the
detection results of the prepared dataset trained on both deep learning models. The
average precision (AP) of both models indicates that the proposed model
outperformed YOLOv3 with 88.03% accuracy in detecting cheating.
Phase-II First Review Department of Computer Science and
Engineering
18
References
[1]H. Cai and I. King, “Education Technology for Online Learning in Times of
Crisis,” in Proceedings of 2020 IEEE International Conference on Teaching,
Assessment, and Learning for Engineering, TALE 2020, Institute of Electrical and
Electronics Engineers Inc., Dec. 2020, pp. 758–763. doi:
10.1109/TALE48869.2020.9368387.
[2]L. Al-Labadi and S. Sant, “Enhance Learning Experience Using Technology In
Class,” J Technol Sci Educ, vol. 11, no. 1, pp. 44–52, Sep. 2021, doi:
10.3926/jotse.1050.
[3]M. Alahmari, “Exploring the Influential Factors Affecting Staff Willingness to
Adopt Augmented Reality,” International Journal of Information and Edu- cation
Technology, vol. 13, no. 7, pp. 1078–1084, Jul. 2023
Phase-II First Review Department of Computer Science and
Engineering
19
References
[4]S. R. Sletten, “Rethinking Assessment: Replacing Traditional Exams with Paper
Reviews,” J Microbiol Biol Educ, vol. 22, no. 2, Sep. 2021, doi:
10.1128/jmbe.00109-21.
[5]É. Cambron-Goulet, J. P. Dumas, É. Bergeron, L. Bergeron, and C. St-Onge,
“Guidelines for Creating Written Clinical Reasoning Exams: Insight from a Delphi
Study,” Health Professions Education, vol. 5, no. 3, pp. 237–247, Sep. 2019, doi:
10.1016/j.hpe.2018.09.001.
[6]R. Ellis, J. Cleland, D. SG. Scrimgeour, A. J. Lee, J. Hines, and P. A. Bren- nan,
“Establishing the predictive validity of the intercollegiate membership of the Royal
Colleges of surgeons written examination: MRCS Part A,” The Surgeon, Aug. 2023,
doi: 10.1016/j.surge.2023.07.004.
Phase-II First Review Department of Computer Science and
Engineering
20
References
[7]M. J. Hoque, Md. Razu, Md. Jashim, and M. Mostafa, “Automation of Tradi-
tional Exam Invigilation using CCTV and Bio-Metric,” International Journal of
Advanced Computer Science and Applications, vol. 11, no. 6, pp. 392–399, 2020,
doi: 10.14569/IJACSA.2020.0110651.
[8]R. E. Ferdig, E. Baumgartner, R. Hartshorne, R. Kaplan-Rakowski, and C. Mouza,
“Teaching, Technology, and Teacher Education During the COVID- 19 Pandemic:
Stories From The Field,” AACE-Association for the Advance- ment of Computing in
Education, pp. 125–128, Jun. 2020.
[9]R. Comas-Forgas, T. Lancaster, A. Calvo-Sastre, and J. Sureda-Negre, “Exam
cheating and academic integrity breaches during the COVID-19 pandemic: An
analysis of internet search activity in Spain,” Heliyon, vol. 7, no. 10, Oct. 2021.
Phase-II First Review Department of Computer Science and
Engineering
21
Paper Publication Status (Phase-I & Phase-II)
PHASE-I
SUSMITA MISHRA, ROOPIKHA S,ROSHINI S,RITIKHA S (2023, November) AUTOMATIC CHEATING DETECTION IN EXAM
HALL.Scopus-Conference.
TITLE:AUTOMATIC CHEATING DETECTION IN EXAM HALL
AUTHOR:SUSMITA MISHRA , ROOPIKHA S, ROSHINI S,
RITIKHA S
CONFERENCE NAME: International Journal of Innovative Science and Modern Engineering (IJISME)
CONFERENCE:Scopus Indexed - Conference
MODEL OF PUBLICATION: OFFLINE
STATUS:ACCEPTED
Phase-II First Review Department of Computer Science and
Engineering
22
Paper Publication Status (PhaseI & PhaseII)
PHASE-II
ROOPIKHA S,ROSHINI S,RITIKHA S
AUTOMATIC CHEATING DETECTION IN EXAM HALL.
TITLE:AUTOMATIC CHEATING DETECTION IN EXAM HALL
MODEL OF PUBLICATION:OFFLINE
For phaseII the publication of journal is ongoing yet not completed
Phase-II First Review Department of Computer Science and
Engineering
23
Thank You
Phase-II First Review Department of Computer Science and
Engineering
24

More Related Content

Similar to Project Phase-II - First Review - PPT Template.pptx

Online Exam Proctoring using Deep Learning
Online Exam Proctoring using Deep LearningOnline Exam Proctoring using Deep Learning
Online Exam Proctoring using Deep LearningIRJET Journal
 
USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...
USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...
USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...ijseajournal
 
A SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATION
A SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATIONA SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATION
A SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATIONIJCSES Journal
 
Archana Verma_Resume
Archana Verma_ResumeArchana Verma_Resume
Archana Verma_ResumeArchana Verma
 
Object oriented sad 6
Object oriented sad 6Object oriented sad 6
Object oriented sad 6Bisrat Girma
 
CRIME EXPLORATION AND FORECAST
CRIME EXPLORATION AND FORECASTCRIME EXPLORATION AND FORECAST
CRIME EXPLORATION AND FORECASTIRJET Journal
 
Smart Surveillance System through Computer Vision
Smart Surveillance System through Computer VisionSmart Surveillance System through Computer Vision
Smart Surveillance System through Computer VisionIRJET Journal
 
IRJET- Lab Monitoring System
IRJET-  	  Lab Monitoring SystemIRJET-  	  Lab Monitoring System
IRJET- Lab Monitoring SystemIRJET Journal
 
IRJET - Face Recognition Door Lock using IoT
IRJET - Face Recognition Door Lock using IoTIRJET - Face Recognition Door Lock using IoT
IRJET - Face Recognition Door Lock using IoTIRJET Journal
 
IRJET- E-Gatepass System
IRJET- E-Gatepass SystemIRJET- E-Gatepass System
IRJET- E-Gatepass SystemIRJET Journal
 
ATM Security System Based on the Video Surveillance Using Neural Networks
ATM Security System Based on the Video Surveillance Using Neural NetworksATM Security System Based on the Video Surveillance Using Neural Networks
ATM Security System Based on the Video Surveillance Using Neural NetworksIRJET Journal
 
IRJET- Automated Student’s Attendance Management using Convolutional Neural N...
IRJET- Automated Student’s Attendance Management using Convolutional Neural N...IRJET- Automated Student’s Attendance Management using Convolutional Neural N...
IRJET- Automated Student’s Attendance Management using Convolutional Neural N...IRJET Journal
 
IRJET - Neural Network based Leaf Disease Detection and Remedy Recommenda...
IRJET -  	  Neural Network based Leaf Disease Detection and Remedy Recommenda...IRJET -  	  Neural Network based Leaf Disease Detection and Remedy Recommenda...
IRJET - Neural Network based Leaf Disease Detection and Remedy Recommenda...IRJET Journal
 
ATM fraud detection system using machine learning algorithms
ATM fraud detection system using machine learning algorithmsATM fraud detection system using machine learning algorithms
ATM fraud detection system using machine learning algorithmsIRJET Journal
 
Systems development
Systems developmentSystems development
Systems developmentElijah Liu
 
MIT521 software testing (2012) v2
MIT521   software testing  (2012) v2MIT521   software testing  (2012) v2
MIT521 software testing (2012) v2Yudep Apoi
 

Similar to Project Phase-II - First Review - PPT Template.pptx (20)

Chapter 2
Chapter 2Chapter 2
Chapter 2
 
Online Exam Proctoring using Deep Learning
Online Exam Proctoring using Deep LearningOnline Exam Proctoring using Deep Learning
Online Exam Proctoring using Deep Learning
 
USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...
USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...
USABILITY EVALUATION OF A CONTROL AND PROGRAMMING ENVIRONMENT FOR PROGRAMMING...
 
A SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATION
A SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATIONA SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATION
A SURVEY OF VIRTUAL PROTOTYPING TECHNIQUES FOR SYSTEM DEVELOPMENT AND VALIDATION
 
Archana Verma_Resume
Archana Verma_ResumeArchana Verma_Resume
Archana Verma_Resume
 
Object oriented sad 6
Object oriented sad 6Object oriented sad 6
Object oriented sad 6
 
Jagadeesh_Resume
Jagadeesh_ResumeJagadeesh_Resume
Jagadeesh_Resume
 
CRIME EXPLORATION AND FORECAST
CRIME EXPLORATION AND FORECASTCRIME EXPLORATION AND FORECAST
CRIME EXPLORATION AND FORECAST
 
Smart Surveillance System through Computer Vision
Smart Surveillance System through Computer VisionSmart Surveillance System through Computer Vision
Smart Surveillance System through Computer Vision
 
IRJET- Lab Monitoring System
IRJET-  	  Lab Monitoring SystemIRJET-  	  Lab Monitoring System
IRJET- Lab Monitoring System
 
IRJET - Face Recognition Door Lock using IoT
IRJET - Face Recognition Door Lock using IoTIRJET - Face Recognition Door Lock using IoT
IRJET - Face Recognition Door Lock using IoT
 
IRJET- E-Gatepass System
IRJET- E-Gatepass SystemIRJET- E-Gatepass System
IRJET- E-Gatepass System
 
ATM Security System Based on the Video Surveillance Using Neural Networks
ATM Security System Based on the Video Surveillance Using Neural NetworksATM Security System Based on the Video Surveillance Using Neural Networks
ATM Security System Based on the Video Surveillance Using Neural Networks
 
IRJET- Automated Student’s Attendance Management using Convolutional Neural N...
IRJET- Automated Student’s Attendance Management using Convolutional Neural N...IRJET- Automated Student’s Attendance Management using Convolutional Neural N...
IRJET- Automated Student’s Attendance Management using Convolutional Neural N...
 
IRJET - Neural Network based Leaf Disease Detection and Remedy Recommenda...
IRJET -  	  Neural Network based Leaf Disease Detection and Remedy Recommenda...IRJET -  	  Neural Network based Leaf Disease Detection and Remedy Recommenda...
IRJET - Neural Network based Leaf Disease Detection and Remedy Recommenda...
 
ATM fraud detection system using machine learning algorithms
ATM fraud detection system using machine learning algorithmsATM fraud detection system using machine learning algorithms
ATM fraud detection system using machine learning algorithms
 
Sample report
Sample reportSample report
Sample report
 
Systems development
Systems developmentSystems development
Systems development
 
SAD_SDLC.pptx
SAD_SDLC.pptxSAD_SDLC.pptx
SAD_SDLC.pptx
 
MIT521 software testing (2012) v2
MIT521   software testing  (2012) v2MIT521   software testing  (2012) v2
MIT521 software testing (2012) v2
 

Recently uploaded

UNIT 4 PTRP final Convergence in probability.pptx
UNIT 4 PTRP final Convergence in probability.pptxUNIT 4 PTRP final Convergence in probability.pptx
UNIT 4 PTRP final Convergence in probability.pptxkalpana413121
 
Maximizing Incident Investigation Efficacy in Oil & Gas: Techniques and Tools
Maximizing Incident Investigation Efficacy in Oil & Gas: Techniques and ToolsMaximizing Incident Investigation Efficacy in Oil & Gas: Techniques and Tools
Maximizing Incident Investigation Efficacy in Oil & Gas: Techniques and Toolssoginsider
 
Passive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.pptPassive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.pptamrabdallah9
 
21P35A0312 Internship eccccccReport.docx
21P35A0312 Internship eccccccReport.docx21P35A0312 Internship eccccccReport.docx
21P35A0312 Internship eccccccReport.docxrahulmanepalli02
 
Augmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptxAugmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptxMustafa Ahmed
 
DBMS-Report on Student management system.pptx
DBMS-Report on Student management system.pptxDBMS-Report on Student management system.pptx
DBMS-Report on Student management system.pptxrajjais1221
 
Databricks Generative AI Fundamentals .pdf
Databricks Generative AI Fundamentals  .pdfDatabricks Generative AI Fundamentals  .pdf
Databricks Generative AI Fundamentals .pdfVinayVadlagattu
 
一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样
一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样
一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样A
 
Working Principle of Echo Sounder and Doppler Effect.pdf
Working Principle of Echo Sounder and Doppler Effect.pdfWorking Principle of Echo Sounder and Doppler Effect.pdf
Working Principle of Echo Sounder and Doppler Effect.pdfSkNahidulIslamShrabo
 
一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书
一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书
一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书c3384a92eb32
 
Theory of Time 2024 (Universal Theory for Everything)
Theory of Time 2024 (Universal Theory for Everything)Theory of Time 2024 (Universal Theory for Everything)
Theory of Time 2024 (Universal Theory for Everything)Ramkumar k
 
Raashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashidFaiyazSheikh
 
Developing a smart system for infant incubators using the internet of things ...
Developing a smart system for infant incubators using the internet of things ...Developing a smart system for infant incubators using the internet of things ...
Developing a smart system for infant incubators using the internet of things ...IJECEIAES
 
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdfInvolute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdfJNTUA
 
Fuzzy logic method-based stress detector with blood pressure and body tempera...
Fuzzy logic method-based stress detector with blood pressure and body tempera...Fuzzy logic method-based stress detector with blood pressure and body tempera...
Fuzzy logic method-based stress detector with blood pressure and body tempera...IJECEIAES
 
01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...
01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...
01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...AshwaniAnuragi1
 
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdflitvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdfAlexander Litvinenko
 
Path loss model, OKUMURA Model, Hata Model
Path loss model, OKUMURA Model, Hata ModelPath loss model, OKUMURA Model, Hata Model
Path loss model, OKUMURA Model, Hata ModelDrAjayKumarYadav4
 
Filters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsFilters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsMathias Magdowski
 

Recently uploaded (20)

UNIT 4 PTRP final Convergence in probability.pptx
UNIT 4 PTRP final Convergence in probability.pptxUNIT 4 PTRP final Convergence in probability.pptx
UNIT 4 PTRP final Convergence in probability.pptx
 
Maximizing Incident Investigation Efficacy in Oil & Gas: Techniques and Tools
Maximizing Incident Investigation Efficacy in Oil & Gas: Techniques and ToolsMaximizing Incident Investigation Efficacy in Oil & Gas: Techniques and Tools
Maximizing Incident Investigation Efficacy in Oil & Gas: Techniques and Tools
 
Passive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.pptPassive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.ppt
 
Signal Processing and Linear System Analysis
Signal Processing and Linear System AnalysisSignal Processing and Linear System Analysis
Signal Processing and Linear System Analysis
 
21P35A0312 Internship eccccccReport.docx
21P35A0312 Internship eccccccReport.docx21P35A0312 Internship eccccccReport.docx
21P35A0312 Internship eccccccReport.docx
 
Augmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptxAugmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptx
 
DBMS-Report on Student management system.pptx
DBMS-Report on Student management system.pptxDBMS-Report on Student management system.pptx
DBMS-Report on Student management system.pptx
 
Databricks Generative AI Fundamentals .pdf
Databricks Generative AI Fundamentals  .pdfDatabricks Generative AI Fundamentals  .pdf
Databricks Generative AI Fundamentals .pdf
 
一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样
一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样
一比一原版(NEU毕业证书)东北大学毕业证成绩单原件一模一样
 
Working Principle of Echo Sounder and Doppler Effect.pdf
Working Principle of Echo Sounder and Doppler Effect.pdfWorking Principle of Echo Sounder and Doppler Effect.pdf
Working Principle of Echo Sounder and Doppler Effect.pdf
 
一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书
一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书
一比一原版(Griffith毕业证书)格里菲斯大学毕业证成绩单学位证书
 
Theory of Time 2024 (Universal Theory for Everything)
Theory of Time 2024 (Universal Theory for Everything)Theory of Time 2024 (Universal Theory for Everything)
Theory of Time 2024 (Universal Theory for Everything)
 
Raashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashid final report on Embedded Systems
Raashid final report on Embedded Systems
 
Developing a smart system for infant incubators using the internet of things ...
Developing a smart system for infant incubators using the internet of things ...Developing a smart system for infant incubators using the internet of things ...
Developing a smart system for infant incubators using the internet of things ...
 
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdfInvolute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
 
Fuzzy logic method-based stress detector with blood pressure and body tempera...
Fuzzy logic method-based stress detector with blood pressure and body tempera...Fuzzy logic method-based stress detector with blood pressure and body tempera...
Fuzzy logic method-based stress detector with blood pressure and body tempera...
 
01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...
01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...
01-vogelsanger-stanag-4178-ed-2-the-new-nato-standard-for-nitrocellulose-test...
 
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdflitvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
 
Path loss model, OKUMURA Model, Hata Model
Path loss model, OKUMURA Model, Hata ModelPath loss model, OKUMURA Model, Hata Model
Path loss model, OKUMURA Model, Hata Model
 
Filters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsFilters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility Applications
 

Project Phase-II - First Review - PPT Template.pptx

  • 1. AUTOMATIC CHEATING DETECTION IN EXAM HALL Mr.PRAVEEN ASST PROFESSOR) ROOPIKHA.S-200701201 ROSHINI.S-200701202 RITHIKA.S-200701198 Department of Computer Science and Engineering CS19811 – Project Phase-II
  • 2. Problem Statement and Motivation The Aim is to develop an automatic cheating detection system using machine learning techniques to identify and prevent cheating behavior during exams conducted in an exam hall environment this system can detect various forms of cheating such as copying from materials, communicating with others using electronic devices or other suspicious behavior while minimizing false positive and negatives cheating during exams determines the integrity of the education system and diminishes the value of academic achievement by using automatic cheating detection system provides more efficient and accurate solution by leveraging advance technologies like computer vision and data analysis this system ensure fairness and maintain the exam protocol. Phase-II First Review Department of Computer Science and Engineering 2
  • 3. Objectives Implementing an automatic cheating detection system in exam halls using machine learning has the following objectives: 1. Enhancing Exam Integrity 2. Minimizing False Positives 3. Ensuring Scalability 4. Reducing Human Errors 5. Achieving Cost and Resource Efficiency These objectives aim to provide a comprehensive and reliable solution to maintain the credibility of exams and uphold the principles of academic honesty. Phase-II First Review Department of Computer Science and Engineering 3
  • 4. Abstract An innovative cheating detection system for exam halls using machine learning is being introduced. By utilizing computer vision and real-time monitoring, the system can identify various cheating behaviors, thus ensuring exam integrity. This model's training minimizes errors and false alerts, optimizing accuracy. Automated alerts help administrators take prompt interventions to uphold fairness and credibility. The system also adheres to privacy regulations, safeguarding student data. This innovative system enhances cheating prevention and maintains the essence of academic honesty. Phase-II First Review Department of Computer Science and Engineering 4
  • 5. System Architecture Phase-II First Review Department of Computer Science and Engineering 5
  • 6. System Architecture Existing System The existing methods for cheating detection in exam halls often rely on manual monitoring by invigilators, which can be time-consuming and prone to human errors.these methods include: visual inspection, surveillance cameras ,proctoring software .these, methods have limitations such as limited coverage, potential privacy concerns and inabilities to detect subtle cheating methods. Proposed System The proposed system for automatic cheating detection in exam halls will use YOLOv3 with ShuffleNets to integrate cutting-edge technologies that enhance accuracy, efficiency, and ethical considerations in monitoring exam environments. Phase-II First Review Department of Computer Science and Engineering 6
  • 7. System Architecture The proposed system is made up of several components that work together to maintain the integrity of an exam hall. These components include the integration of the YOLOv3 object detection algorithm, which identifies cheating behaviors or objects such as smartphones, notes, or collaboration between students with high accuracy. Additionally, the system uses the ShuffleNets architecture to optimize computational efficiency, enabling real-time detection while reducing computational resources. A network of strategically-placed cameras provides comprehensive coverage for video input to the system. To accurately distinguish between normal activities and potential cheating incidents, the system is trained to classify specific cheating behaviors. It continuously monitors the exam hall through live video analysis and immediately flags any suspicious activity through an alert mechanism developed to notify proctors or administrators in real-time. Phase-II First Review Department of Computer Science and Engineering 7
  • 8. System architecture System Operation: The system is designed to adapt to varying lighting conditions, camera angles, or exam hall layouts for consistent and reliable performance. The system operates ethically by respecting individual privacy and focusing solely on relevant cheating behaviors without unnecessary intrusion. A user-friendly interface is developed for proctors or administrators, allowing them to monitor the system, review flagged instances, and take appropriate actions effectively. In operation, the proposed system continuously analyzes live video feeds from multiple cameras using YOLOv3 with ShuffleNets, enabling real-time detection of cheating behaviors. Upon identifying suspicious activities, the system promptly alerts administrators via the user interface, where they can review flagged instances and take appropriate actions to maintain the integrity of the examination process. Phase-II First Review Department of Computer Science and Engineering 8
  • 9. List of Modules 1. Clearly define the objectives and goals of your project: What exactly do you want to achieve with automatic cheating detection using YOLOv5 2. Identify the Stakeholders: Identify all the key stakeholders, including end-users, project sponsors, and anyone else who will be impacted by or involved in the project. 3. Gather Requirements: Collect and document the specific requirements for the cheating detection system. These may include data sources, performance metrics, and user interface requirements. Phase-II First Review Department of Computer Science and Engineering 9
  • 10. List Of Modules 4. Data Collection and Annotation: Determine what data you will need for training and testing your YOLOv5 model. This includes images or videos of test-taking environments. Annotate the data to identify cheating activities. 5. Model Selection: Decide on the use of YOLOv5 as the object detection framework. Assess its suitability for your specific project requirements. 6. Hardware and Software Requirements: Identify the hardware and software infrastructure needed to run YOLOv5 efficiently. Consider GPU/CPU resources, storage, and software dependencies. Phase-II First Review Department of Computer Science and Engineering 10
  • 11. List Of Modules 7. Team Formation: Assemble a team with the required skills, such as machine learning engineers, data scientists, and software developers. 8. Project Timeline: Create a detailed project timeline that outlines milestones, deadlines, and deliverables. 10. Data Preprocessing: Clean, preprocess, and augment the data as necessary for training and testing the YOLOv5 model. Ensure data privacy and security considerations are met. Phase-II First Review Department of Computer Science and Engineering 11
  • 12. Functional Description for each modules with DFD and Activity Diagram Phase-II First Review Department of Computer Science and Engineering 12
  • 13. Implementation/Results of Module Project Description for the First Module: Objective: The primary objective of the first module is to develop an automatic cheating detection system using YOLOv3 object detection with ShuffleNets architecture, specifically designed for monitoring and preventing cheating behaviors in an exam hall or similar testing environments. Key Features of Functionalities: 1. Real-time Object Detection: Our system uses YOLOv3 (You Only Look Once) to provide accurate object detection in real-time. Phase-II First Review Department of Computer Science and Engineering 13
  • 14. Implementation/Results of Module 2. Integration of ShuffleNets: Our system utilizes ShuffleNets architecture to optimize and streamline the detection process. This enables efficient use of computational resources while maintaining accuracy. 3. Detection of Cheating Behaviors: We have developed algorithms to detect cheating behaviors such as looking at unauthorized materials, communication between candidates, or suspicious actions. 4. Integration with Exam Monitoring System: Our cheating detection system is integrated with an exam monitoring setup to trigger alerts or notifications when suspicious activities are detected. Phase-II First Review Department of Computer Science and Engineering 14
  • 15. Implementation/Results of Module 5. Data Logging and Reporting: Our system captures and logs instances of suspicious behavior for further review and analysis by authorized personnel. This module is designed to cater to the needs of educational institutions, examination centers, or any organization that conducts exams where maintaining the integrity of the testing process is of utmost importance. The primary goal of this module is to assist invigilators or administrators in managing and monitoring exam halls to ensure that examinations are conducted fairly and by the established guidelines. Technologies and Tools: Phase-II First Review Department of Computer Science and Engineering 15
  • 16. Implementation/Results of Module We have implemented the YOLOv3 deep learning model to detect objects in real time. To optimize computational efficiency without compromising accuracy, we have employed the ShuffleNets architecture. Python is the primary programming language we use for implementation. Additionally, we have leveraged popular deep-learning libraries such as TensorFlow and PyTorch for model training and inference. The estimated timeline for the completion of the first module is as follows: Research and Design: 1 month, Model Development and Integration: 2 months, Testing and Refinement: 1.5 months, Integration with Exam Monitoring System: 1 month. Phase-II First Review Department of Computer Science and Engineering 16
  • 17. Implementation/Results of Module The initial module's outcomes aim for robust cheating detection in exam halls using YOLOv3 with ShuffleNets. Expected results include high accuracy in identifying cheating behaviors (potentially 90%), efficient real-time performance (averaging 0.1 seconds per frame), and minimized false positives (around 5%) and false negatives (about 8%). Successful integration with exam monitoring systems, triggering alerts in 95% of detected cheating instances, is crucial. Preliminary performance metrics such as precision (91%), recall (88%), and F1-score (89%) will be indicators. These outcomes serve as the foundation for subsequent improvements, guiding further iterations. Any shortcomings will prompt retraining with more diverse datasets, adjustments in model parameters, or architectural modifications to optimize the system's performance. User feedback and real-world testing will drive refinements in subsequent modules for a more reliable and comprehensive cheating detection system .Phase-II First Review Department of Computer Science and Engineering 17
  • 18. Pending Modules & Conclusion The proposed model was implemented using Python with Tensor Flow on a core i7 system with Nvidia 1080 ti, and trained with a learning rate of 0.0002, batch size of 1, for 100 epochs, and momenta of 0.091. The training started with a loss of 4.02323, which gradually reduced to 0.0023. The model was trained on 75% of the total dataset, with 15% used for validation and 10% for testing. The proposed model achieved an accuracy of 88.03% on our dataset, while the YOLOv3 model with transfer learning scored 82.06% accuracy. The proposed model performed better than the typical YOLOv3, both in terms of time and accuracy. Below Table shows the detection results of the prepared dataset trained on both deep learning models. The average precision (AP) of both models indicates that the proposed model outperformed YOLOv3 with 88.03% accuracy in detecting cheating. Phase-II First Review Department of Computer Science and Engineering 18
  • 19. References [1]H. Cai and I. King, “Education Technology for Online Learning in Times of Crisis,” in Proceedings of 2020 IEEE International Conference on Teaching, Assessment, and Learning for Engineering, TALE 2020, Institute of Electrical and Electronics Engineers Inc., Dec. 2020, pp. 758–763. doi: 10.1109/TALE48869.2020.9368387. [2]L. Al-Labadi and S. Sant, “Enhance Learning Experience Using Technology In Class,” J Technol Sci Educ, vol. 11, no. 1, pp. 44–52, Sep. 2021, doi: 10.3926/jotse.1050. [3]M. Alahmari, “Exploring the Influential Factors Affecting Staff Willingness to Adopt Augmented Reality,” International Journal of Information and Edu- cation Technology, vol. 13, no. 7, pp. 1078–1084, Jul. 2023 Phase-II First Review Department of Computer Science and Engineering 19
  • 20. References [4]S. R. Sletten, “Rethinking Assessment: Replacing Traditional Exams with Paper Reviews,” J Microbiol Biol Educ, vol. 22, no. 2, Sep. 2021, doi: 10.1128/jmbe.00109-21. [5]É. Cambron-Goulet, J. P. Dumas, É. Bergeron, L. Bergeron, and C. St-Onge, “Guidelines for Creating Written Clinical Reasoning Exams: Insight from a Delphi Study,” Health Professions Education, vol. 5, no. 3, pp. 237–247, Sep. 2019, doi: 10.1016/j.hpe.2018.09.001. [6]R. Ellis, J. Cleland, D. SG. Scrimgeour, A. J. Lee, J. Hines, and P. A. Bren- nan, “Establishing the predictive validity of the intercollegiate membership of the Royal Colleges of surgeons written examination: MRCS Part A,” The Surgeon, Aug. 2023, doi: 10.1016/j.surge.2023.07.004. Phase-II First Review Department of Computer Science and Engineering 20
  • 21. References [7]M. J. Hoque, Md. Razu, Md. Jashim, and M. Mostafa, “Automation of Tradi- tional Exam Invigilation using CCTV and Bio-Metric,” International Journal of Advanced Computer Science and Applications, vol. 11, no. 6, pp. 392–399, 2020, doi: 10.14569/IJACSA.2020.0110651. [8]R. E. Ferdig, E. Baumgartner, R. Hartshorne, R. Kaplan-Rakowski, and C. Mouza, “Teaching, Technology, and Teacher Education During the COVID- 19 Pandemic: Stories From The Field,” AACE-Association for the Advance- ment of Computing in Education, pp. 125–128, Jun. 2020. [9]R. Comas-Forgas, T. Lancaster, A. Calvo-Sastre, and J. Sureda-Negre, “Exam cheating and academic integrity breaches during the COVID-19 pandemic: An analysis of internet search activity in Spain,” Heliyon, vol. 7, no. 10, Oct. 2021. Phase-II First Review Department of Computer Science and Engineering 21
  • 22. Paper Publication Status (Phase-I & Phase-II) PHASE-I SUSMITA MISHRA, ROOPIKHA S,ROSHINI S,RITIKHA S (2023, November) AUTOMATIC CHEATING DETECTION IN EXAM HALL.Scopus-Conference. TITLE:AUTOMATIC CHEATING DETECTION IN EXAM HALL AUTHOR:SUSMITA MISHRA , ROOPIKHA S, ROSHINI S, RITIKHA S CONFERENCE NAME: International Journal of Innovative Science and Modern Engineering (IJISME) CONFERENCE:Scopus Indexed - Conference MODEL OF PUBLICATION: OFFLINE STATUS:ACCEPTED Phase-II First Review Department of Computer Science and Engineering 22
  • 23. Paper Publication Status (PhaseI & PhaseII) PHASE-II ROOPIKHA S,ROSHINI S,RITIKHA S AUTOMATIC CHEATING DETECTION IN EXAM HALL. TITLE:AUTOMATIC CHEATING DETECTION IN EXAM HALL MODEL OF PUBLICATION:OFFLINE For phaseII the publication of journal is ongoing yet not completed Phase-II First Review Department of Computer Science and Engineering 23
  • 24. Thank You Phase-II First Review Department of Computer Science and Engineering 24