Project 1 : Covid-19 detection using lung sounds
In this project I made a Covid-19 detector which uses Lung sounds to analyse and predict the chances of the disease. I used HTML, CSS, JAVASCRIPT and Google Teachable Machine to make the classifier and the website . Check the website out here :
https://mayukhmalidas-covid19lungtest.netlify.app/
PLEASE OPEN THE ABOVE LINK IN PC FOR BEST RESULTS AND TURN ON THE MIC IN YOUR PC
The You-tube playlist from where I extracted the audio is :
https://www.youtube.com/playlist…
Any suggestions is highly appreciated
PS: The classifier is very easy to build, Just put the audio files in Google Teachable Machine and then build a classifier. Then import the Jason file in JavaScript, Cook it up with some Html and CSS and vola - the website will be complete.
You will find the code for the above project here :
https://github.com/Smartmayukh/Covid-19-detection-using-lung-sounds
Project 2 : Fake News during Covid-19
My second project is " Fake News during Covid-19 " . Initially you can see nodes in different shades of violet. These are the nodes in a social media site. You can start the fake news on any node, by clicking on it. You can also adjust the consciousness of the people to find out how much influence it has on spreading of Fake news. Finally there are two bars where you can see how many people ignored the message and how many forwarded it.
You can check out the project here :
https://mayukhmalidascovid19.netlify.app/
Open this link in computer for better results.
You can check out our YouTube video about Fake news here :
You will find the code for the above project here :
https://github.com/Smartmayukh/Covid-19-fake-news-
3rd and 4th project : BD CAM 1.0 and Mask detector
There are two videos below. The second one is a simple mask detector which any one can make within few minutes without any coding. We are, I think, way past that phase for mask detection. The first video focuses on an app which I am developing known as BD CAM. It is an app that will detect if one has Covid-19 or not, by facial detection. Doctors have already invented similar apps for Tuberculosis, Parkinson's disease etc. As we know in India there is and will be a shortage of Test kits. So we need an alternative, available to all. This ML Model has pictures of different symptoms as classes. It finds correlation amongst them. But it is not efficient enough. It requires more data and more epochs. After this model is over, the plan is to find a correlation between all these percentages - multiplying all those percentages by certain constant values and adding them and if it is over a certain threshold value then one has Covid-19. Then it will be deployed for mobile devices. But it will take time to prepare the exact model with all the correlations . It is an open source project so anyone can collaborate. Thank you. Let's not waste whatever time we have at our disposal and fight this pandemic.
I made both of the above projects using Google Teachable Machines . The link for Teachable Machines is :
https://teachablemachine.withgoogle.com/
5th Project: Hand-wash timer
This is a simple hand-wash timer made using Arduino Uno, Ultrasonic sensor and Servo motor. The Ultrasonic sensor detects if someone has come to wash their hands. The servo motor receives this signal from the Ultrasonic sensor and starts ticking for 20 seconds. Such a visual display will help to motivate people to wash their hands for 20 seconds.
6th project : FLATTEN THE CURVE INDIA
This is a glimpse of my model ' FLATTEN THE CURVE INDIA '. The video is in 16X speed. This ML Model made in SPYDER is fully customizable. It includes total population, percentage of infected people at the beginning, radius of transmission, probability of transmission, percentage of people in quarantine and time taken to recover. The model will approximately calculate the time when the curve will start to flatten. It will also give approximately accurate answers for large demographic countries like India considering approximate population densities. I will upload the codes by tomorrow and also a video about how to use this to optimize and reduce the length of Covid-19 in India. You will surprised to find out that lock-down doesn't play that much of a role in this optimization process as you think. BUT in densely populated countries like India, lock-down is absolutely necessary to prevent community transmission, but that's not enough to fight this pandemic. The improvement of recovery rate, reduction of radius of contact and some other are the key features which will truly help us.
Final project : Covid-19 detection using Chest X ray
Currently working on Covid-19 detection using Chest X ray of patients. I have already gathered approved databases of X- ray, the Covid X-ray database is from GitHub and the normal and the pneumonia ones are from Kaggle.
I have made all the above projects under 2 weeks , working about 2 hours each day behind them , but now I am focusing on to my studies . I will make more new and unique projects soon, when time permits me to do so. Thanks a lot for reading this blog about my projects . Please leave your valuable feedback in the comment section below .
THANK YOU
Comments