I'm new to deep learning and I'm trying out a challenge that where I need to classify species of whale. The requirement is that I need to match 5 individual ids with an image evaluated with map@5
I'm having trouble with how I should process the data. The csv has a structure like this for the training set:
The test set does not include species name and for each image in the test set I may predict up to 5 individual ids. For every new individual in the test set a prediction of new_individual should be made.
Right now I'm just trying to figure out If I need to even include the species column while processing the data on the test csv and images.
I plan on using tensorflow during this.
EDIT: Individual IDs correlate to species. There are 30 species and 15587 individual ids. There is a total of 51033 images.