Search papers, labs, and topics across Lattice.
3
0
4
MLLMs can be distilled into lightweight arbiters that dramatically improve the robustness of composed image retrieval by disentangling noisy training signals.
HABIT achieves superior image retrieval performance by simulating human habit formation, effectively tackling the Noise Triplet Correspondence problem that plagues traditional methods.
Recovering static 3D scenes from monocular video with dynamic objects gets a boost: GA-GS leverages diffusion models to inpaint occluded regions, outperforming existing methods, especially in scenarios with large-scale occlusions.
