# ImageNet X

https://facebookresearch.github.io/imagenetx/site/home

ImageNetの手動の要因設計によるアノテーションの追加

2200個の現存するモデルに適用

(1) architecture – e.g. transformer vs.　convolutional –

(2) learning paradigm – e.g. supervised vs. self-supervised –, and 

(3) training procedures –e.g. data augmentation.

Regardless of these choices, we find models have consistent failure modes across ImageNet-X categories

# 無職期間の過ごし方

https://tech-camp.in/note/pickup/101333/