https://github.com/borda/kaggle_covid-detection

Identify and localize COVID-19 abnormalities on chest radiographs.
https://github.com/borda/kaggle_covid-detection

covid-19 image-classification object-detection

Last synced: 19 days ago
JSON representation

Identify and localize COVID-19 abnormalities on chest radiographs.

• Host: GitHub
• URL: https://github.com/borda/kaggle_covid-detection
• Owner: Borda
• License: mit
• Created: 2021-08-03T06:59:13.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2022-01-04T13:21:37.000Z (about 3 years ago)
• Last Synced: 2024-12-27T16:40:49.076Z (28 days ago)
• Topics: covid-19, image-classification, object-detection
• Language: Jupyter Notebook
• Homepage: https://borda.github.io/kaggle_COVID-detection
• Size: 1.16 MB
• Stars: 5
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE
  • Codeowners: .github/CODEOWNERS

Awesome Lists containing this project

README

        
# Kaggle: [COVID-19 Detection](https://www.kaggle.com/c/siim-covid19-detection)

[![CI complete testing](https://github.com/Borda/kaggle_COVID-detection/actions/workflows/ci_testing.yml/badge.svg?branch=main&event=push)](https://github.com/Borda/kaggle_COVID-detection/actions/workflows/ci_testing.yml)

[![Code formatting](https://github.com/Borda/kaggle_COVID-detection/actions/workflows/code-format.yml/badge.svg?branch=main&event=push)](https://github.com/Borda/kaggle_COVID-detection/actions/workflows/code-format.yml)

[![codecov](https://codecov.io/gh/Borda/kaggle_COVID-detection/branch/main/graph/badge.svg)](https://codecov.io/gh/Borda/kaggle_COVID-detection)

[![pre-commit.ci status](https://results.pre-commit.ci/badge/github/Borda/kaggle_COVID-detection/main.svg)](https://results.pre-commit.ci/latest/github/Borda/kaggle_COVID-detection/main)

In this competition, you’ll identify and localize COVID-19 abnormalities on chest radiographs.

In particular, you'll categorize the radiographs as negative for pneumonia or typical, indeterminate, or atypical for COVID-19.

Organizers provided dataset - imaging data and annotations from a group of radiologists.

![Sample images](./assets/image-class-samples.jpg)

In other words the task to solve is image classification accompanied by attention

- user shall classify patient COVID19 situation and also tell why think so, what are the regions in the scan that makes him think this case is positive.

![Label distribution](./assets/labels-pie.png)

## Experimentation

### install this tooling

A simple way how to use this basic functions:

```bash

! pip install https://github.com/Borda/kaggle_COVID-detection/archive/main.zip

```

### see local notebook

- [COVID19 detection with Flash ⚡](notebooks/COVID-detection-with-Lightning-Flash.ipynb)

### run notebooks in Kaggle

- [COVID19 detection with Flash ⚡](https://www.kaggle.com/jirkaborovec/covid-detection-with-lightning-flash)

- [COVID19 detection - predictions](https://www.kaggle.com/jirkaborovec/covid-detection-with-lightning-flash-predictions)

### some results

Training progress with ResNet50 with training  for 50 epochs:

![Training process](./assets/logging-loss.png)

![Training process](./assets/logging-metric.png)