search

yaoyao-liu / class-incremental-learning

PyTorch implementation of AANets (CVPR 2021) and Mnemonics Training (CVPR 2020 Oral)
https://class-il.mpi-inf.mpg.de
MIT License
461 stars 71 forks source link

Difference between subset_ImageNet vs miniImageNet #26

Open InugYoon opened 3 years ago

InugYoon commented 3 years ago

Hello, thank you for sharing great project. I wonder, what is difference between subset-ImageNet and miniImageNet?

They both are subset of ImageNet and have 100 classses. So, would there be any reason for using subset-ImageNet instead of miniImageNet?

yaoyao-liu commented 3 years ago

The miniImageNet was proposed for few-shot classification in this paper.

The ImageNet-Subset was proposed for class-incremental learning in this paper.

So, miniImageNet and ImageNet-Subset are two different datasets for different tasks. If we hope to compare our method with existing work, we need to use the corresponding datasets.

InugYoon commented 3 years ago

Oh, okay thank you!!