ArticleJournal of vision2021

Use of superordinate labels yields more robust and human-like visual representations in convolutional neural networks.

Seoyoung Ahn et al.PubMed ↗Full text ↗Publisher ↗

No numbers read from the abstract.