After visualizing the dataset, Data Explorer provides us with several explorations options:
We can click on each dot, displaying the images that are very similar to that one by using the KNN option. For example, here we can see similar cars.
Clicking on the ‘Random’ button provides us with a random sampling of the entire database. WE can even see which images have been chosen to be displayed and further processed.
Next, by using the ‘Manual’ option, we can choose each image independently, such as the vehicle from the example beforehand or choosing another one from each one of the clusters to be added to the displayed list.