arXiv - CSCL: "Coarse-to-Fine Contrastive Learning in Image-Text…" - Qoto Mastodon

arXiv - CSCL @arxiv_cscl@qoto.org

Coarse-to-Fine Contrastive Learning in Image-Text-Graph Space for Improved Vision-Language Compositionality. (arXiv:2305.13812v3 [cs.CL] UPDATED)

http://arxiv.org/abs/2305.13812 #arXiv #NLProc

Oct 26, 2023, 03:17 · · arxiv-cscl · · ·

Sign in to participate in the conversation