Search papers, labs, and topics across Lattice.
D assets, while faithfully retaining the visual characteristics of the original input. Figure 1: Fast and versatile
2
0
4
Edit 3D assets with text prompts while actually preserving the original object's unchanged parts, thanks to a new masking strategy and training dataset.
Noisy labels tank dynamic pruning performance, but AlignPrune's loss-trajectory alignment recovers up to 6.3% accuracy without architecture or training changes.
