Fengyuan Yang

Hi, I think the explanation you are looking for can be found at the end of 'Training Procedure' in '3.3. Framework' of our paper: > More specifically, for each episode,...

Thank you for bringing this to our attention, I've updated the [cub.zip](https://drive.google.com/file/d/13qzg_yv_uRP5vUprpzLwfYPyi0_R_guo/view?usp=sharing) which contains the split file of CUB.

Hi, the key is to generate 3 pair (data-label pair) of files from your dataset. It doesn't matter what format they are in. For example, I simply generate six .npy...

Hi, CUB-FS images are cropped with given bounding boxes following the previous works [DeepEMD ](https://github.com/icoz69/DeepEMD) and [FEAT](https://github.com/Sha-Lab/FEAT).

Hello, the [models link](https://drive.google.com/file/d/1eS49e5Wt5gXnMM7TLso1eS2B04kX8oLh/view?usp=sharing) is uploaded in the README.md.

Hi, since MiniImagenet is a derivative of ImageNet dataset and ImageNet is organized according to the [WordNet ](https://wordnet.princeton.edu/) hierarchy, the WordNet ID (or WordNet synset) of each MiniImagenet class can...

Hi, the _cubfs_attributes.npy_ is directly the 312 attributes of each bird category given by the original CUB dataset.