‘25
SPRINTATHON ‘25
A SPRINT TO REVOLUTION
1 Problem Statement Number- PSE405
2 Problem Statement – Smart Attendance and Engagement Monitoring System
3 Domain - SMART EDUCATION
4 Team Details - Recursius
5 Team Members Name - LOKESH SELVAM , VISHNU MUTHIAH, SHAAN NARENDRAN
College Name - COLLEGE OF ENGINEERING GUINDY, ANNA UNIVERSITY
6
‘25
PROPOSED SOLUTION- ARGUS
Argus is an intelligent, real-time platform that leverages computer vision to automate student attendance and
analyze classroom engagement. By interpreting non-invasive body language cues, the system provides educators
with actionable insights into student attentiveness, helping them optimize teaching methodologies and enhance the
overall learning experience.
What makes us unique:
• Overall engagement information: Giving information not just on individuals, but on the classroom as a whole helps
pinpoint the issues that affect the majority of students and not just a select few.
• Real-time analysis: Live graph plotting delivers immediate feedback, allowing teachers to change their methods
during the lesson.
• Unique Composite monitoring of engagement: The model relies on multiple factors in conjunction to give definitive,
intuitive results.
Addressing the issue
• Manual attendance is replaced with automated facial recognition, simplifying the process considerably.
• When monitoring engagement, it uses the most indicative and basal body language to articulate its results,
providing reliable and meaningful data.
• It also supplies educators with objective, data-driven feedback on student focus and overall lecture effectiveness.
‘25
TECHNICAL APPROACH
Mock UI:
Working Prototype: Technologies Used:
Database – File System (Image-based) Backend – Python
Frontend – Streamlit
https://github.com/ShaanNarendran/Sprintathon-Attendance-Tracker
AI Models – MediaPipe, OpenCV DNN
Computation – NumPy, SciPy
‘25
FEASIBILITY AND VIABILITY
FEASIBILITY: Strategies to Overcome:
• Technological Feasibility: The project is built on Ethical: Implement a "Privacy-by-Design" architecture,
mature, high-performance, and open-source libraries like
anonymize all analytics data, and build a transparent consent
Google's MediaPipe, drastically reducing development
model.
time and risk.
• Economic Feasibility: The reliance on standard Technical: Employ frame-processing optimization techniques
hardware (webcams, PCs) eliminates the need for and design the system for scalable deployment on cloud GPUs.
expensive, specialized equipment, making it highly
accessible to a wide range of educational institutions.
User Adoption: Design an intuitive dashboard with built-in
CHALLENGES: tutorials, and avoid grading or otherwise academically
impacting the students through the tool’s use.
• Ethics & Privacy: Gaining user trust to securely and
ethically handle what would otherwise be considered Data Accuracy & Algorithmic Bias: Continuously refine the AI
sensitive and intrusive student biometric data.
model with diverse, real-world classroom data to distinguish
• Scalability: Ensuring smooth real-time performance in between genuine disengagement and activities like note-taking
a crowded classroom setting where several students or deep thought. Actively audit and mitigate algorithmic biases
need to be monitored simultaneously by training the system on a dataset representative of diverse
• User Adoption: Overcoming the perception of a ethnicities, genders, and learning accommodations to ensure
surveillance tool as opposed to an educational one, and equitable and fair analysis for all students.
ensuring teachers can easily interpret the data.
‘25
IMPACT & SOCIAL RELEVANCE
Potential Impact on the Target Audience
● For Educators: The platform is designed to empower educators with objective, real-time data on
classroom dynamics. This can help refine teaching strategies, identify challenging topics, and automate the
administrative task of taking attendance.
● For Students: The intended impact is an improved learning environment that is more responsive to
student focus levels. The system also provides a mechanism for educators to identify and offer timely
support to students who may be consistently disengaged.
Alignment with Broader Social Needs & Global Goals
● Social Needs: The project addresses the increasing need for data-informed educational practices,
particularly with the growth of hybrid and large-scale learning environments where individual student
engagement is difficult to assess.
● Global Goals: The solution provides a tool to enhance teaching effectiveness and foster more inclusive
classroom engagement, regardless of learning methods, language, or other such barriers.
‘25
MARKET DEMAND ANALYSIS
● Target Audience
The primary audience includes educators in K-12 and Higher Education institutions, as well as corporate training
departments. The solution is particularly relevant for large classes or hybrid learning environments.
● Specific Need
There is a clear market need for tools that move beyond simple attendance to provide objective, actionable
insights into student engagement levels.
● Unique Value Proposition
This solution combines automated attendance, multi-metric passive engagement analysis (focus, alertness,
fatigue), and active participation tracking (hand raising) into a single, synthesized dashboard.
● Market Potential and Supporting Trends
The global EdTech market is experiencing substantial growth, with a clear trend towards the adoption of AI-driven
analytics. The increasing prevalence of hybrid and remote learning models further solidifies the demand for
effective engagement solutions.
‘25
RESEARCH AND REFERENCES
Details / Links of the reference and research work
Research paper on Face Recognition based Attendance System :
https://www.ijert.org/research/face-recognition-based-attendance-system-IJERTV9IS060615.pdf
Paper on Facial emotion recognition using deep learning review and insights:
https://www.sciencedirect.com/science/article/pii/S1877050920318019
Eye Aspect Ratio for Real-Time Drowsiness Detection to Improve Driver Safety:
https://www.researchgate.net/publication/364268066_Eye_Aspect_Ratio_for_Real-Time_Drowsin
ess_Detection_to_Improve_Driver_Safety
