Dear Colleagues,

It’s been roughly four months since the last update to our COVID-Net initiative.

Today, we are extremely proud to announce an evolutionary leap to the effort by way of COVID-Net CT-2.

Image for post
Image for post

COVID-Net CT-2 was built using a number of large and diverse datasets that we painstakingly created over several months with researchers from the University of Waterloo. The largest of the datasets consists of a multinational cohort of over 4500 patients across 15 countries (with over 200,000 CT slices) and has been fully curated and standardized. …


Explainability tells us how AI works, but says nothing about when it can be trusted. This limitation must be confronted.

Image for post
Image for post

Earlier this year, in our XAI primer, we introduced the concept of explainability by way of a provocative thought experiment; namely, by asking readers to consider a major scientific advancement and remove the foundational knowledge that enabled it:

At Kitty Hawk in 1904, the Wright brothers chance upon on a strange and unwieldy metallic structure. Its design and aerodynamic properties are a mystery, but through much trial-and-error, they’re flying…

In such a scenario, we argued, the euphoria of the moment…


Dear Colleagues,

It’s been a while since our last update regarding our COVID-Net initiative.

Today, we are extremely proud to announce a new and significant update to the project: COVIDNet-S.

COVIDNet-S is a suite of deep learning models designed to assess the disease severity of COVID-19.

As viral procedures like RT-PCR and antibody tests cannot diagnose the severity of COVID-19 for a given patient, chest X-Rays have routinely been leveraged to assist health care professionals in tracking the progression of the disease.

To this end, COVIDNet-S can quantitatively score the geographic and opacity extent in a patient’s lungs by analyzing…


Dear Colleague,

The response to our release of COVID-Net three months ago continues to inspire us.

Here are the latest updates with the project:

1.) Today we are proud to announce the release of COVIDNet-CT, a family of neural networks to detect COVID-19 by means of Computed Tomography (CT) scans. Over 100,000 CT images from ~1500 patients were leveraged to build the model. The source code, trained models, data preparation scripts, training/inference/evaluation scripts, and instructions can be accessed at this GitHub repository, which will be regularly updated with better models and additions to data.

In keeping with the spirit of…


“Imagination is the highest form of Research” — Albert Einstein

On March 22nd our team announced the availability of COVID-Net (latest updates here), a neural network for COVID-19 detection using chest radiography (X-Rays). Our decision to open source the model, dataset and project source code, attracted an overwhelming response from researchers around the world and a significant amount of media coverage (here, here and here).

From the outset, we were clear that COVID-Net was intended as a complementary tool to assist clinicians in rapidly screening for the virus and not a replacement for the definitive but slower and more complex…


Dear Colleague,

The response to our release of COVID-Net continues to move and inspire us.

Here are the latest updates with the project:

  1. Thanks to the continued support of the global community, the total number of COVID-19 positive cases in the COVIDx dataset has grown significantly, with the test benchmark set now more than 3x larger than the last release. This makes COVIDx one of the largest open-access benchmark datasets in terms of number of COVID-19 positive cases. Instructions and scripts on generating this latest collection are available on our GitHub repo. …


Dear Colleague,

The response to our release of COVID-Net continues to move and inspire us.

Here are the latest updates with the project:

1.) Thanks to support of the research community, the number of COVID-19 positive cases in the training and test portions of the COVIDx dataset has increased by 2x and 3x, respectively. Instructions and scripts on generating this latest collection can are available on our GitHub repo. We’d especially like to thank Figure 1 for their assistance.

2.) We’ve released two models based on the dataset above: COVIDNet-CXR-Large and COVIDNet-CXR-Small. The former provides higher detection sensitivity for COVID-19…


Note: since this post was published there’ve been a number of developments to the project. See here for the latest.

Dear Colleague,

The response to our release of COVID-Net one week ago has been overwhelming.

Here are the latest updates with the project:

1.) Thanks to the support and feedback from the research community, the COVIDx dataset now consists of 16,756 chest X-Rays across 13,645 patient cases. Instructions and scripts on generating this enhanced collection can be found on our GitHub repo.

We’d like to thank the following groups for making their data available as it accelerates efforts such as…


Note: since this post was published there’ve been a number of developments to the project. See here for the latest.

Dear Colleague,

The global crisis brought on by COVID-19 has affected us all.

Like many businesses, we’ve been grappling with how to best deploy our skills in service of the present crisis.

To this end, we have collaborated with researchers at the University of Waterloo’s VIP Lab to develop COVID-Net: a convolutional neural network for COVID-19 detection via chest radiography. …


AI’s ‘black box’ may be the greatest business and societal risk of our time. Here’s how we bring it to light.

Image for post
Image for post

by Sheldon Fernandez

As was illustrated in Part I of this piece, existing approaches to explainability are limited in important ways. Some are indirect, while others lack stability. Some require subjective interpretation, while others lack verifiability.

In other words, to the extent that AI is black box, current solutions are but tiny pinholes in its protective coating, facilitating small but insufficient rays of sunlight through deep learning’s dark exterior.

How can we more fully illuminate the abyss?

At the…

Sheldon Fernandez

CEO, DarwinAI

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store