COVID-VIT: Classification of Covid-19 from CT chest images based on vision transformer models

Last update: Dec 30, 2022

Related tags

Overview

COVID-ViT

COVID-VIT: Classification of Covid-19 from CT chest images based on vision transformer models

This code is to response to te MIA-COV19 competition on classification of covid from non-covid chest volumetric CT datasets.

Pre-trained models for ViT and DenseNet can be download from https://drive.google.com/drive/folders/1nBI02F-8Y0hFeN10CMu9Svj4xRViN5xt?usp=sharing.

Both 2D and 3D versions of training and test code are provided. It appears classificaiton based on 2D slices performs better. The final score is subject based, i.e. for a dataset, if more than 25% or more slices are classfied as COVID, then this subject has COVID. Otherwise, the patient in concern will be classified as normal. This threshold (e.g 25%) can be determined from validation stage.

The ViT is heavily based on vit-pytorch at https://github.com/lucidrains/vit-pytorch and is in the form of both notebook and python.

The DenseNet-CT is built upon https://github.com/UCSD-AI4H/COVID-CT.

More details are at the paper ar Arxiv (https://arxiv.org/) with the following information: "Xiaohong Gao, Yu Qian, Alice Gao, COVID-VIT: Classification of Covid-19 from CT chest images based on vision transformer models"

COVID-VIT: Classification of Covid-19 from CT chest images based on vision transformer models

Related tags

Overview

COVID-ViT

Owner

An algorithmic trading bot that learns and adapts to new data and evolving markets using Financial Python Programming and Machine Learning.

Sky Computing: Accelerating Geo-distributed Computing in Federated Learning

This is a collection of our NAS and Vision Transformer work.

An interactive DNN Model deployed on web that predicts the chance of heart failure for a patient with an accuracy of 98%

Official PyTorch implementation of the paper "Deep Constrained Least Squares for Blind Image Super-Resolution", CVPR 2022.

Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data

U-Net: Convolutional Networks for Biomedical Image Segmentation

Best practices for segmentation of the corporate network of any company

PyTorch-based framework for Deep Hedging

IsoGCN code for ICLR2021

Official PyTorch implementation of "Improving Face Recognition with Large AgeGaps by Learning to Distinguish Children" (BMVC 2021)

2021-AIAC-QQ-Browser-Hyperparameter-Optimization-Rank6

Deep Learning for Time Series Classification

ADB-IP-ROTATION - Use your mobile phone to gain a temporary IP address using ADB and data tethering

TorchDistiller - a collection of the open source pytorch code for knowledge distillation, especially for the perception tasks, including semantic segmentation, depth estimation, object detection and instance segmentation.

Laplacian Score-regularized Concrete Autoencoders

Convnet transfer - Code for paper How transferable are features in deep neural networks?

GAN-generated image detection based on CNNs

Python Implementation of algorithms in Graph Mining, e.g., Recommendation, Collaborative Filtering, Community Detection, Spectral Clustering, Modularity Maximization, co-authorship networks.

Codebase to experiment with a hybrid Transformer that combines conditional sequence generation with regression