Compression then Matching: An Efficient Pre-training Paradigm for Multimodal Embedding

Add code
Nov 11, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: