REAL TIME FACE MASK DETECTION
USING DEEP LEARNING
A Mini Project Report submitted in partial fulfillment of the requirements for the award of
the degree of
BACHELOR OF TECHNOLOGY
IN
COMPUTER SCIENCE AND ENGINEERING (AI & ML)
Submitted by
Student Name (USN No)
Under the guidance of
Mrs. XYZ
Assistant Professor, Department of CSE (AI & ML)
DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING
XYZ COLLEGE OF ENGINEERING AND TECHNOLOGY
(Affiliated to ABC University, Approved by AICTE and Accredited by NBA & NAAC)
City, State - Zip Code
2024 – 2025
CERTIFICATE
This is to certify that the project report entitled “REAL TIME FACE MASK DETECTION
USING DEEP LEARNING” submitted by [Student Names & USNs] in partial fulfillment of the
requirements for the award of the degree of Bachelor of Technology in Computer Science
and Engineering (AI & ML) is a record of bonafide work carried out under our guidance and
supervision.
Project Guide,
Mrs. XYZ
Assistant Professor
Department of CSE (AI & ML), XYZ College
ACKNOWLEDGEMENT
We are grateful to our guide Mrs. XYZ for her guidance and support throughout the project.
We extend our thanks to the Head of Department, Principal, faculty members, and our peers
for their encouragement and assistance. Finally, we thank our families for their constant
support.
PROJECT STUDENTS:
Student 1 (USN)
Student 2 (USN)
Student 3 (USN)
Student 4 (USN)
ABSTRACT
The real-time face mask detection system aims to identify individuals who are or aren’t
wearing a face mask in public areas using deep learning. The system uses Convolutional
Neural Networks (CNNs) with OpenCV and TensorFlow for efficient image processing and
classification. This solution is important in public health monitoring and safety enforcement
during pandemics such as COVID-19.
CONTENTS
CHAPTER 1 – INTRODUCTION
CHAPTER 2 – FACE MASK DETECTION
CHAPTER 3 – DEEP LEARNING
CHAPTER 4 – CONVOLUTION NEURAL NETWORKS
CHAPTER 5 – IMPLEMENTATION
CHAPTER 6 – RESULTS AND DISCUSSION
CHAPTER 7 – CONCLUSION
REFERENCES
CHAPTER 1: INTRODUCTION
The COVID-19 pandemic created a global need for monitoring and ensuring face mask
compliance in public areas. This project proposes an AI-powered system that uses deep
learning to detect whether people are wearing masks in real time. The motivation stems
from public health needs and the importance of automation in surveillance.
CHAPTER 2: FACE MASK DETECTION
Face mask detection refers to identifying whether a person in an image or video is wearing
a face mask. This technology gained momentum during the COVID-19 pandemic when
governments and organizations enforced the use of face masks in public areas to mitigate
virus transmission. Face mask detection is typically achieved using machine learning and
deep learning algorithms that classify images based on the presence or absence of a mask.
Key processes include:
- Face Detection: Locating the presence of a human face within an image using techniques
such as Haar Cascades or deep learning-based detectors like MTCNN.
- Mask Classification: Once a face is detected, a classifier determines if the face is wearing a
mask or not. CNN models trained on large datasets of masked and unmasked faces are used.
- Real-Time Processing: For practical use in surveillance, the detection must be performed
in real time using efficient models and optimization techniques.
Applications include:
- Automated monitoring systems in hospitals, airports, offices, and schools
- Public transport surveillance for safety compliance
- Integration into attendance systems or access control mechanisms
CHAPTER 3: DEEP LEARNING
Deep learning is a subset of machine learning that uses neural networks with many layers
to model complex patterns in data. Inspired by the structure and functioning of the human
brain, deep learning has revolutionized fields such as computer vision, natural language
processing, and speech recognition.
A typical deep learning model consists of:
- Input Layer: Receives raw data such as images or audio.
- Hidden Layers: Multiple layers of neurons that extract features and transform the input
into meaningful patterns.
- Output Layer: Produces predictions or classifications.
Neural networks learn by adjusting weights through backpropagation. Each neuron
performs a weighted sum of its inputs and passes the result through an activation function.
Common functions include:
- ReLU (Rectified Linear Unit)
- Sigmoid
- Tanh
The training process involves minimizing a loss function such as binary cross-entropy using
optimizers like SGD or Adam. Deep learning frameworks like TensorFlow and PyTorch have
made it easier to build and deploy models.
In face mask detection, deep learning enables high accuracy by automatically learning
spatial features such as the presence of a mask across different angles and lighting
conditions.
