The document presents a project dissertation on 'Face Detection using Machine Learning' by students at Samarth Group of Institutions, focusing on developing a machine learning-based approach for effective face detection in challenging scenarios. It outlines the problem of face detection, reviews existing literature, and proposes an architecture that promises improved performance over traditional methods. The work emphasizes the use of feature extraction and classification techniques for enhanced detection accuracy in images and videos.

1. Samarth Group of Institutions College of Engineering,
Belhe, Junnar, Pune – 412 410
Department of Computer Engineering
A Project Disseration Stage-I on
“ Face Detection using Machine Learning ”
Presented By
1. Shinde Harsha K. ( 54 )
2. Auti Siddhi K. ( 66 )
3. Gadhave Snehal G. ( 68 )
4. Gawali Kiran D. ( 69 )
5. Veer Shivani U. ( 58 )
S.E. – P.B.L. (Computer Engineering)
Under the guidance of
Prof. Shelke S. D.

2. Outline
• Abstract
• Introduction
• Literature Survey
• Problem Statement
• System Architecture Diagram
• Reference

3. Abstract
Face detection is a critical task in a variety of
applications, including security and surveillance,
human-computer interaction, and social media. In this
work, we propose a machine learning-based approach
to face detection that aims to improve performance in
difficult or unconstrained scenarios. Our approach
utilizes a combination of feature extraction and
classification techniques to accurately detect and
locate faces in images or video. We evaluate our
system on a standard dataset of images and video, and
demonstrate improved performance compared to
baseline approaches. The proposed system is robust,
efficient, and has the potential to significantly impact
the field of face detection.

4. Introduction
Face detection is a key problem in computer vision,
with numerous applications. This work proposes a
machine learning-based approach to improve face
detection in challenging or unconstrained scenarios.
Our method uses a combination of feature extraction
and classification techniques to detect and locate
faces in images or video. We evaluate the
performance of our system on a standard dataset and
show improved results compared to baseline
approaches. This approach is efficient and has
potential to significantly impact the field of face
detection.

5. Literature
Survey
Face detection is a well-studied problem in computer vision, with a wide
range of applications including security and surveillance, human-computer
interaction, and social media. Over the past few decades, a variety of
approaches have been proposed to solve the face detection problem, including
traditional computer vision techniques and machine learning algorithms. Early
face detection approaches relied on hand-crafted features and simple
classifiers, such as Haar cascades [1] and local binary patterns [2]. These
methods were effective for many applications but struggled to handle
variations in pose, lighting, and facial expression. More recent approaches
have employed machine learning techniques to improve face detection
performance. These methods often use deep neural networks (DNNs) to learn
features directly from data, achieving state-of-the-art results on a variety of
benchmarks [3, 4, 5]. However, these methods can be computationally
expensive and may not be suitable for real-time or resource-constrained
applications. There has also been a focus on developing methods that are
robust to poor image quality or partial occlusion, which can be common in
real-world scenarios [6, 7]. These approaches often incorporate additional
information, such as facial landmarks or depth maps, to improve performance.
Overall, the field of face detection has made significant progress in recent
years, but there is still room for improvement, particularly in challenging or
unconstrained scenarios.

6. Problem
Statement
•The problem of face detection involves detecting and
locating human faces in digital images or video. This
is a challenging problem due to the wide range of
variations in appearance that can occur due to factors
such as pose, lighting, and facial expression. In order
to address this problem, we propose to use machine
learning algorithms to automatically detect and locate
faces in images or video.

7. System
Architecture
Diagram

8. Reference
[1] P. Viola and M. J. Jones. Rapid object detection using a boosted cascade
of simple features. In Proceedings of the 2001 IEEE Computer Society
Conference on Computer Vision and Pattern Recognition (CVPR), volume 1,
pages I-511-I-518 vol.1, 2001.
[2] T. Ojala, M. Pietikäinen, and T. Mäenpää. Multiresolution gray-scale and
rotation invariant texture classification with local binary patterns. IEEE
Transactions on Pattern Analysis and Machine Intelligence, 24(7):971-987,
2002.
[3] K. Zhang, Z. Zhang, Z. Li, and Y. Qiao. Joint face detection and
alignment using multitask cascaded convolutional networks. IEEE Signal
Processing Letters, 23(10):1499-1503, 2016.
[4] Y. Liu, Z. Li, and Y. Qiao. Recognizing faces using deep convolutional
neural networks. In Proceedings of the 2015 IEEE International Conference
on Computer Vision (ICCV), pages 3730-3738, 2015.
[5] S. Zhang, Z. Li, and Y. Qiao. Joint face detection and alignment with
multitask learning. In Proceedings of the 2017 IEEE International
Conference on Computer Vision (ICCV), pages 1433-1441, 2017.
[6] Y. Liu, Z. Li, and Y. Qiao. Learning multi-view face detection using deep
convolutional neural networks. In Proceedings of the 2016 European
Conference on Computer Vision (ECCV), pages 707-723, 2016.
[7] S. Li, X. Li, and Y. Qiao. A convolutional neural network cascade for
face detection. In Proceedings of the 2016 IEEE Conference on Computer
Vision and Pattern Recognition (CVPR), pages 5325-5334, 2016.

9. Thank You