Abstract: In this paper, we look at cross-domain few-shot clas-sification which presents the challenging task of learning new classes in previously unseen domains with few la-belled examples. Existing ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results