The SCNN- xBlock model - Simple Covid Neural Network with x blocks - was designed with the objective of being simple and efficient at analyzing chest x-rays. The architecture is composed of the following:
Link to access the model weights: https://drive.google.com/drive/folders/1-3NdJBOLZVh9sbTva8srQLtdrCfLhcKY?usp=sharing
Link to access the model weights: https://drive.google.com/drive/folders/1oVuymbP1irJu65j8bDx-sjjbMQW0V_tY?usp=sharing
This architecture was developed with simplicity in mind, given the not so good results on the bigger models. As such, starting really simple, we hoped the increase of CNN's blocks would improve performance to a certain point, staying at 2 blocks. Maybe with a bigger dataset, the results would be better.