Dataset > Modality > Images > miniImageNet

miniImageNet

Introduced by Vinyals et al. in Matching Networks for One Shot Learning

The miniImageNet dataset contains 100 classes randomly chosen from ImageNet ILSVRC-2012 challenge with 600 images of size 84×84 pixels per class. It is split into 64 base classes, 16 validation classes and 20 novel classes

Source: Leveraging the Feature Distribution in Transfer-based Few-Shot Learning

Samples

Modalities

Languages