Fetching the paper…

LocTex: Learning Data-Efficient Visual Representations from Localized Textual Supervision · Around