The code is written in pythonon on kaggle!
The model employed to solve this problem is a Convolutional Neural Network (CNN), which achieved 100% accuracy on the test dataset. To enhance its performance, I leveraged the architecture and pre-trained weights of the VGG16 model. By incorporating aspects of the VGG16 model, the CNN was able to achieve optimal results for the task!
Comments are made in swedish as i did this on my spare time before my master.