Covid-19 Test AI (Deep Learning – NNs) Software

First of all, hello everyone!
I can’t wait to tell you about my discovery, which has the distinction of being the only one in the world, about Covid-19. First of all, I would like to tell you a little about myself.
I am a 20 year old deep learning engineer. I have been interested in deep learning and algorithms for about 7 years. I have been continuing my Machine Learning and Deep Learning studies for a long time. So at the age of 13, I was curious about this field. Even though I think that I have successfully passed this life full of financial and hardware inadequacies and various obstacles, I am still a deep learning engineer who has difficulties in realizing my projects in terms of hardware and material. Despite all these insufficient opportunities, on October 29, 2021, around 07.30 pm UK time, I offered my Covid-19 Artificial Intelligence Test software to all humanity for free on GitHub. My goal is to change the world and equip every corner of the world with technological instruments!
Now I will give you brief information about the project.

Project Purpose – Covid-19 Test AI

The aim of the project is to be an artificial intelligence system based on reducing the current false rate of coronavirus tests in the world, saving time and reducing costs. As a matter of fact, PCR tests are carried out for a fee. Maybe you don’t pay for PCR testing, but the government buys this technology on your behalf. At this point, I encoded various neuronal networks that detect the X-RAY X-ray system to increase the well-being of humanity. And I created them with multiple layers. I tried and tried! Then I ran it on the tagged data. In addition, I abstracted the lungs with the segmentation algorithm I created. In this way, we can detect whether the virus is present.

COVID-19 Test AI Software
COVID-19 Test AI Software Cover Image

Artificial Intelligence Accuracy Rate – Covid-19 Test AI

I saw the accuracy rate of the software I created was around 99% in 200 epochs. The loss rate was around 2% (0.02). That was an incredible score! But I shared my version of 102-110 epochs with humanity. Here, the accuracy is 96.86% and the loss rate is 0.0901%. The NNs I created trained and used the data wonderfully! They learned them very well! Well done to them 🙂

12/12 [==============================] – 55s 4s/step – loss: 0.0901 – accuracy: 0.9686
Model: “model”

Developed by Emirhan BULUT

About Data

I showed the address where I got the data as a link on GitHub. I would like to share with you the following about the data:

Class (Covid Test)Images (numbers)Women (numbers) Men (numbers) Severities (numbers)
negative426239187
positive426190236Normal-PCR+: 76, Mild: 100, Moderate: 171, Severe: 79
Table Source

I performed a segmentation show with the labeled parts of the data. It is clear that this test is positive or negative for covid-19!

Materials Used in the Covid-19 Test AI Project

I provided all the materials in the project with my own means. These materials are:

  • My own computer (worth $200),
  • 10Mbps internet,
  • PyCharm is an IDE program.

Libraries I used in the project are:

  • Tensorflow,
  • keras,
  • pixellib,
  • OpenCV,
  • PIL.
tensorflow-keras-PIL-pixellib-opencv-cv2-library
Libraries: Tensorflow, Keras, PIL, Pixellib, OpenCV

In Summary… For Covid-19 Test AI (Covid-19 Test Artificial Intelligence)

I am giving away my Covid-19 Artificial Intelligence test software (Covid-19 Test AI) to all humanity for free. This software was created by me for free for the benefit of all humanity. All software codes and models are uniquely designed by me. Algorithms have been designed by me with precision in order to provide high accuracy and benefit.

My goal is to change the world, and I’m very close to that!

Kind regards,

Emirhan BULUT
Deep Learning and Artificial Intelligence Engineer
Covid-19 Test AI Inventor

Project Link: https://github.com/emirhanai/COVID-19-Test-AI-Software

My GitHub: https://github.com/emirhanai

My LinkedIn: https://www.linkedin.com/in/artificialintelligencebulut/

My Kaggle: https://www.kaggle.com/emirhanai

Leave a Comment