Search papers, labs, and topics across Lattice.
1
0
3
A single, lightweight model can now outperform specialized video grounding models and massive MLLMs, thanks to a clever query unification strategy that enables effective cross-dataset learning.