sayyedabrarakhtar.com.np
theme
~/portfolio

// ACADEMIC CONTENT PORTAL · Computer Science

Thesis Topic Ideas: E-Attendance System Using Facial Recognition

Level: BachelorDifficulty: Intermediate★ Popular Choice

🎓 Academic Journey Progression Map

1. Introduction & Problem Statement

Overview: Design and implement an automated student/employee attendance system using computer vision facial detection and recognition algorithms linked to a web dashboard.

Background Context (Nepal): Manual paper attendance in Nepali colleges and offices is time-consuming and vulnerable to proxy marking. Automated face recognition streamlines attendance logging.

2. Research Objectives

  • Train lightweight face detection and recognition model (OpenCV / dlib / FaceNet)
  • Build real-time webcam video feed capture interface
  • Integrate automated database logging matching detected face embeddings with student IDs
  • Develop web portal dashboard for faculty attendance reporting and CSV export
  • Achieve 92%+ recognition accuracy under variable indoor lighting conditions

3. Proposed Methodology

  1. Collect face image dataset of 50 student volunteers with 10 poses/lighting variations each
  2. Extract 128-d face embeddings using dlib / OpenCV facial landmark detector
  3. Implement k-NN or SVM classifier matching live frame embeddings to database records
  4. Build Next.js / React management dashboard backed by Node.js and MongoDB
  5. Evaluate accuracy, false acceptance rate (FAR), and latency per recognition frame

$ Worked Example / Sample Scenario

Sample Scenario: A student walks past the computer lab camera. The dlib pipeline detects facial landmarks, computes 128-d embeddings, matches with ID #2081-CS-042 in 0.4 seconds, and marks attendance on the college portal.

4. Thesis Chapter-by-Chapter Outline

Chapter 1: IntroductionTU/KU/NEB standard

Background, problem statement, research questions, objectives, scope, and significance

Chapter 2: Literature ReviewTU/KU/NEB standard

Computer vision architectures, HOG vs CNN face detection, and embedding matching models

Chapter 3: System Design & MethodologyTU/KU/NEB standard

System architecture, ER diagrams, embedding pipeline, and dataset preparation

Chapter 4: Implementation & ResultsTU/KU/NEB standard

Accuracy evaluation, confusion matrix, frame rate benchmarks, and web portal screenshots

Chapter 5: Conclusion & Future WorkTU/KU/NEB standard

Summary of findings, limitations under extreme low light, and mobile app export roadmap

5. Recommended Tools & Technologies

To implement the practical, technical, or analytical portions of this thesis topic, the following software tools, libraries, or APIs are recommended:

PythonOpenCVdlib / FaceNetNext.jsNode.jsMongoDB

6. Core References & Academic Sources

  • [1]IEEE Xplore / Google Scholar — Computer Vision Facial Recognition Papers <!-- VERIFY: needs confirmation -->
  • [2]NepJOL — Applied AI in academic administrative automation

7. Frequently Asked Questions (FAQs)

Q: Does face recognition work with eyeglasses or face masks?

Using facial landmark embeddings focused on eye-bridge and forehead keypoints allows partial recognition under minor occlusion.

Related Academic Topics & Student Tools

available for workKathmandu, Nepal 🇳🇵contact@sayyedabrarakhtar.com.np