cewein / bee-on-artifical-flower Goto Github PK

As of now the dataset is dictionary stored in a .npy file (numpy format). But the dataset doesn't properly interfere with some neural network.

On the other hand pascalVOC and the COCO dataset are the state of the art in dataset. Therefore, formating the dataset to one or both of these format whould help a lot in the the long run.

here is the link to both dataset :

https://cocodataset.org/#format-data
http://host.robots.ox.ac.uk/pascal/VOC/

Please, reference this issue when working on the dataset changes.