English | 简体中文 | Tiếng việt

Tracking Students

This is a part of the 2023 startup project by a group consisting of four members: Dat Nguyen Tien, Nguyen Truong Phuc, Nguyen Thanh Tu, and Luyen Thanh Binh, under the guidance of PhD. Pham Minh Chuan. The project deals with the issue of managing students during examinations. Specifically, this section of code manages students within the examination room. Its task is to supervise, monitor, and detect any cheating candidates through facial recognition attendance. The detailed code segment is provided below.

Contents

1. Introduction
2. Results
3. Installation
4. Quick Start Overview
5. Structures
6. Send Us Feedback!
7. Thanks
8. License

Introduction

Face Recognition

Facial recognition is a rapidly developing technology that has attracted much attention in recent years. This technology involves automatically identifying and verifying individuals based on their facial features. It relies on collecting, analyzing, and comparing unique facial characteristics, such as the distance between the eyes and the shape of the nose.

With the use of deep learning algorithms, facial recognition systems have become highly accurate and are applied in various fields, including security, access control, and even on social media platforms. Implementing this system in examinations and education is an essential necessity. This system helps control and identify candidates within the examination room. It extracts facial information and, upon completion, exports it to Excel for cross-referencing when needed.

Counting Students

Alongside facial recognition technology, the technology for monitoring students during examinations has also seen significant advancements. This technology employs one of the most modern AI models available today. The system automatically captures and sends images to Telegram every 5 minutes, enabling educators to monitor student entries and exits while recording the timing of these movements with real-time results every 10 seconds.

By combining both machine learning and deep learning algorithms, the system for monitoring candidates is becoming increasingly accurate. Its application is gaining trust among people and expanding widely in both daily life and professional settings.

Tracking People

Another field that is rapidly developing and showing even greater potential in the future is the detection of cheating among candidates using AI. When operational, this system marks and captures images if it detects any anomalies or violations, sending them to Telegram for verification of the misconduct. The reliability of this system is currently at a credible level and is undergoing further development and testing.

This deep learning-based system is being applied in developed countries worldwide such as the UK, France, the USA, and various Asian countries like Japan, South Korea, among others. It is being implemented in collaboration with examination invigilators to achieve the highest effectiveness and ensure the utmost fairness in examinations.

Results

Counting Students (using Yolov8 )

^{Testing with 12422TN class on Yolov8}

Cheating Recognition ( using OpenPose)

^{Testing with 12422TN class on OpenPose}

Installation

With Python Base

Requirements python >= 3.7

1. Install dependences library

pip install -r setup.txt

2. Install dependences files

• After you must run file_requirements.py or

python file_requirements.py

3. Install dependences files with other steps ( Optional )

• If you step 2 not successfully you can download weights from Google Drive
• Move folder graph_models downloaded to Pose\graph_models and copy folder you downloaded tosrc\Pose\graph_models\
• Paste file in folder Face-Recognition\Models you downloaded to Models in main

With Anaconda

1. Install dependences library
   • You can load dependences library with env.yaml file.
   • You can find env.yaml file in folder Anaconda
2. Install dependences files After you must run file_requirements.py or

python file_requirements.py

3. Install dependences files with other steps ( Optional )

• If you step 2 not successfully you can download weights from Google Drive
• Move folder graph_models downloaded to Pose\graph_models and copy folder you downloaded tosrc\Pose\graph_models\
• Paste all file in folder Face-Recognition\Models you downloaded to Models in main

With Docker

You can build docker images with my docker file.

1. Build docker images In folder main of this project open command prompt and run

docker build -t [names_you_choose] .
Example: docker build -t nguyendat135/trackingstudents .

Quick Start Overview

With Python Base Environments and Anaconda Environment

1. Quick Run

• You can run this file app.py to start this project.
• Input if login username:abc and password:abc to login
• The IP and Port you can access is http://localhost:8080/ or with other laptop or smartphone in same network is http://192.168.1.44:8080

2. To trainning model you read Hướng dẫn sử dụng in tab Tổng quan

With Docker

1. Quick Run You access this project with this command

docker run -p 8080:8080 [name_you_choose in Installation]
Example: docker run -p 8080:8080 nguyendat135/trackingstudents

2. After you can access it with http://localhost:8080/ or with other laptop or smartphone in same network is http://192.168.1.44:8080
3. To trainning model you read Hướng dẫn sử dụng in tab Tổng quan

Structures

Tracking_Students
+---.idea
¦   +---inspectionProfiles
+---Action
¦   +---training
¦   +---__pycache__
+---Auth
¦   +---__pycache__
+---Dataset
¦   +---FaceData
+---graph_models
¦   +---mobilenet_thin
¦   +---VGG_origin
+---Models
+---Pose
¦   +---graph_models
¦   ¦   +---mobilenet_thin
¦   ¦   +---VGG_origin
¦   +---__pycache__
+---profile_detection
¦   +---haarcascades
¦   +---__pycache__
+---src
¦   +---Action
¦   ¦   +---training
¦   ¦   +---__pycache__
¦   +---align
¦   ¦   +---__pycache__
¦   +---Auth
¦   ¦   +---__pycache__
¦   +---FCRN
¦   +---generative
¦   ¦   +---models
¦   +---models
¦   +---Pose
¦   ¦   +---graph_models
¦   ¦   ¦   +---mobilenet_thin
¦   ¦   ¦   +---VGG_origin
¦   ¦   +---__pycache__
¦   +---QSTP
¦   +---SoNguoi
¦   +---Tracking
¦   ¦   +---deep_sort
¦   ¦   ¦   +---__pycache__
¦   ¦   +---graph_model
¦   ¦   +---__pycache__
¦   +---ViPham
¦   +---__pycache__
+---test
+---test_out
+---Tracking
¦   +---deep_sort
¦   ¦   +---__pycache__
¦   +---graph_model
¦   +---__pycache__
+---trained
+---ViPham
+---__pycache__

Send Us FeedBack

Our project is open source for research purposes, and we want to improve it! So let us know (create a new GitHub issue or pull request, email us, etc.) if you...

1. Find/fix any bug (in functionality or speed) or know how to speed up or improve any part of Students Tracking.
2. Want to add/show some cool functionality/demo/project made on top of Students Tracking. We can add your project link to your Issue

Thanks

Thank you for the guidance of PhD. Minh Chuan Pham in the process of creating this project, as well as the evaluation board consisting of PhD. Quoc Viet Hoang and PhD. Dinh Chien Nguyen, who helped us improve the results and provided feedback for this project.

License

This project is freely available for free non-commercial use. If it useful you can give 1 star. Thanks for using.