Quick-Draw-Subset

A subset of the famous Google "Quick Draw" dataset ( https://github.com/googlecreativelab/quickdraw-dataset )

The original dataset is very noisy and big, this dataset was created for images similarity algorithms or for evaluation of these.

The dataset

• classes num: 182
• images num: 18.200 (100 images per class)
• images: 128x128 / RGB

A non noisy dataset

• the data was downloaded and sorted by similarity with the "Shape Context" algorithm, and then checked manually (yes... manually 😓)
• too similar and too noisy classes are filtered and deleted (this is the reason why it contains only 182 classes, and not 345)

Can a Backpack looks like a dog?

The answer is yes.

This is only one example of noise removed in this subset.

Download

https://www.kaggle.com/emiliuz/quick-draw-subset