I am looking for the large Japanese Animation Dataset online that is labeled. It is better to have it at least not only the face. I see many on Kaggle, but they are not labeled or face only.
Have you seen this?
The dataset (which they link) has over 140,000 faces
Thank You for your help. But these are not labeled, and anything that is totally generated by Gan is so unrealistic or unstable.
What sort of labels or problem type are you looking for?
What Iβm asking is how do you want the labels to be set up? What is your βlabelβ in this case? Is it the anime? The name of the character? Or are you wanting a diverse unlabeled dataset to do colorization with (to which no label is per say needed)
Also if you read the link to the dataset I posted above, their dataset was not generated anime faces. It was anime faces that exist⦠to which they then generated faces off of
IIRC such a dataset doesnβt exist to the extent ImageNet is, so you may want to start by making a small one. Perhaps fine-graining and labeling one of the anime picture datasets that exists. Others may have better luck or know more than I and could chime in 
1 Like
Useful repo: https://github.com/deeppomf/DeepLearningAnimePapers
1 Like
ι‘εηγδΈι’¨ε€γγ£γγ’γγ‘η»εγ«ε€γγ¦γΏγ¦ο½
γγΌγΏγ»γγγγγε ΄εγ―γγγγ試γγ¦γΏγΎγ~γ