EN HR
Covid Virus - CDC/Unsplash
Covid Virus - CDC/Unsplash

The COVID-19 disease has changed the lives of billions of people around the globe in a matter of weeks. Every step of putting the COVID-19 pandemic under control will contribute to prevention of deaths, make everyday life easier, and put us on a path to brighter future.

Velebit AI team has decided to help the efforts against COVID-19 in every way our expertise makes that possible.

In recent weeks our team has produced an interactive web page to track COVID-19 statistics in Croatia, we have open-sourced an enhanced model for detecting COVID-19 from X-ray images of patients’ lungs, and we are actively working on new projects. Sharing hardware resources for computing research to discover new therapeutic opportunities for COVID-19 has also been among our efforts.

During these tough times, collaboration among interdisciplinary teams is crucial for tackling the virus problem, and therefore, we invite all teams and research groups that see a cooperation potential to contact us. Together, we can perform a lot more in less time!

Covid 19 CRO Visualizations
Covid 19 CRO Visualizations

On the graphs relevant for Croatia, the most important dates of the social distancing measures entering into force are highlighted:

  • March 13 - closure of schools
  • March 17 - ban on social gatherings and closure of cafe bars
  • March 22 - earthquake in Zagreb
  • March 23 - suspension on public transport and ban on leaving the place of residence

COVID-19 HR Dashboard

Interactive data visualization COVID-19 Croatia

Successful fight against COVID-19 begins with a clear presentation of quality data. The interactive website with data visualizations enables monitoring of daily updated statistics and visualizations related to COVID-19 in Croatia, and neighbouring countries. The website aggregates all relevant and available official data sources which are then visualized in a user-friendly manner.

The system is currently monitoring and displaying data on the number of confirmed cases, infected, deceased and cured patients and as well as the number of tested cases. The data is displayed as the number of new cases on a daily basis and the total number of COVID-19 cases. On users' request, visualisations such as the daily growth rate and the daily number of samples tested per confirmed case are also being displayed.

Data on the number of confirmed cases and their daily increase are also given on the county level. The website also monitors and displays basic information about the neighbouring countries and the most affected European countries.

We plan on upgrading the COVID-19 visualizations and your feedback is more than welcome.

If you're interested in the integration of our graphs, visualizations or data into your own website, feel free to contact us.

Signs of pneumonia in a chest X-ray
Signs of pneumonia in a chest X-ray

AI model for detecting COVID-19 from X-ray lung images

Inspired by the recent paper COVID-Net: A Tailored Deep Convolutional Neural Network Design for detection of COVID-19 Cases from Chest Radiography Images, MIT Technology Review article, and Tensorflow implementation of this paper, we are open-sourcing the upgraded Pytorch implementation.

The model, named COVID-Next, features an architecture that builds upon the famous ResNext50 architecture, which has around 5x fewer parameters than the original COVID-Net and achieves comparable performance.

The project aims to encourage broad adoption and contribution which is why the project is licensed under the permissive MIT License.

Tensorflow and Pytorch are two major deep learning frameworks and our motivation was to give the Pytorch research community the same starting ground Tensorflow already has when it comes to AI COVID-19 research. As the authors from the paper have already mentioned, this model still doesn't offer production-ready performance. The key issue that needs to be resolved is the number of COVID-19 images as the number of such images is currently not diverse and large enough to provide representative prediction results end-users could expect in the production system.

SARS-CoV-2 main viral protease
SARS-CoV-2 main viral protease

Sharing distributed computing resources

Project Folding@home provides one of the most simple ways one can help the research community in the battle against COVID-19. This is why we at Velebit AI have decided to chip in with our own computing resources.

Folding@home (FAH or F@h) is a distributed computing project for simulating protein dynamics, including the process of protein folding and the movements of proteins implicated in a variety of diseases. Insights from this data are helping scientists to better understand biology and provide new opportunities for developing therapeutics. People behind the project have tirelessly been working to add new computing jobs related to SARS-CoV-2 virus to discover new therapeutic opportunities for COVID-19. For more details about specific computing jobs check out their blog post.

If you have free CPU or GPU computing resources at your hand, they can be used to help tackle SARS-CoV-2 related tasks. Folding@home client can be run either by Linux, Windows or macOS. The installation process is simple and you can find it in the official installation guides.

We build AI for your needs.

Meet our highly experienced team, who loves to build AI and design its surroundings to incorporate it in your business. Find out how much you can benefit from our fair and open approach.

Contact Us

Members of