SOCIAL DISTANCING
DETECTION
AND
ENFORCEMENT TOOL
PROJECT REVIEW 1
Introduction to Innovative Projects
(PHY1901)
By
Bimal Parajuli (20BDS0405)
Bijan Shrestha (20BCE2904)
Lijah Gongal (20BCE29XX)
Shreema Gautam (20BCE29XX)
SLOT: TA1
FACULTY: Prof. Bhaskar Sen Gupta
1
ABSTRACT
Social Distancing – the term that had taken the world by storm and transformed the
way we lived during the pandemic and thereafter. Social distancing has become a
mantra around the world, transcending languages and cultures.
This way of living has been forced upon us by the fastest growing pandemic the
world has ever seen – COVID-19. As per the World Health Organization (WHO),
COVID-19 has so far infected almost 600 million people and claimed over 6.5
million lives globally. Every country in the world had been affected by the deadly
virus.
The biggest cause of concern was that COVID-19 could spread from person to
person through contact or if you’re within proximity of an infected person. Given
how densely populated some areas are, this has been quite a challenge.
Social distancing is an important way to slow down the spread of infectious
diseases. People are asked to limit their interactions with each other, reducing the
chances of the disease being spread with physical or close contact.
2
MAIN INSPIRATION OF THE PROJECT
Many nations had locked down to contain the epidemic, and the government
compelled the inhabitants to remain at home during the pandemic. The
governmental services such as the Centers for Disease Control and Prevention
(CDC) had to clarify that it is by avoiding intimate contacts with others that we may
most effectively halt the spread of Covid-19. Citizens across the world adopted
physical distance to flatten the curve on the Covid-19 epidemic. To reduce the
effects of this coronavirus pandemic, social distance and self-isolation were
identified as the most effective techniques to interrupt the chain of infections when
economic activity is resumed.
MAIN OBJECTIVE OF THE PROJECT
The objective of this project is to make it easier to enforce social distancing by
employing a deep learning model to detect violations of social distance in
workplaces and public spaces. Various approaches for object detection exist in the
fields of machine learning and computer vision. These approaches may also be
used to determine people's social distances
3
STATISTICAL ANALYSIS
As we all know, maintaining social distance while wearing proper masks is the
most fundamental way to reduce the risk of getting or spreading the COVID-19
virus. A statistical figure below shows the importance of social distancing and
how frequent contact with people can increase the spread.
5
METHODOLOGY
COMPONENTS TO BE USED:
1. Faster R-CNN
Fast R-CNN is an object detector that was developed solely by Ross Girshick, a
Facebook AI researcher, and a former Microsoft Researcher. Fast R-CNN overcomes
several issues in R-CNN. As its name suggests, one advantage of the Fast R-CNN
over R-CNN is its speed.
A Faster R-CNN object detection network is composed of a feature extraction
network which is typically a pretrained CNN, similar to what we had used for its
6
predecessor. This is then followed by two subnetworks which are trainable. The
first is a Region Proposal Network (RPN), which is, as its name suggests, used to
generate object proposals and the second is used to predict the actual class of the
object. So, the primary differentiator for Faster R-CNN is the RPN which is
inserted after the last convolutional layer. This is trained to produce region
proposals directly without the need for any external mechanism like Selective
Search. After this we use ROI pooling and an upstream classifier and bounding
box regressor like Fast R- CNN.
2. OpenCV
OpenCV (Open-Source Computer Vision Library) is an open source computer
vision and machine learning software library. OpenCV was built to provide a
common infrastructure for computer vision applications and to accelerate the
use of machine perception in the commercial products. Being a BSD-licensed
product, OpenCV makes it easy for businesses to utilize and modify the code.
Initially, the main aim of creating OpenCV was real-time applications for
computational efficiency. Since 2011, OpenCV also offers GPU acceleration
for real-time operations. Upon integration with other libraries, such as
NumPy, Python can process the OpenCV array structure for analysis.
Identifying image patterns and its several features needs use of vector space
and carrying out mathematical operations on these features.
7
3. Detectron 2
Facebook AI Research (FAIR) came up with this advanced library, which gave
amazing results on object detection and segmentation problems. Detectron2 is
based upon the maskrcnn benchmark. Its implementation is in PyTorch. It
requires CUDA due to the heavy computations involved.
It supports multiple tasks such as bounding box detection, instance segmentation,
keypoint detection, densepose detection, and so on. It provides pre-trained
models which you can easily load and use it on new images.
20
Tentative Block Diagram for the Project
20
Feasibility analysis for the project.
Social relevance and usability of the project.
20
Expected Project Timeline
20
CONCLUSIONS
20
