Search papers, labs, and topics across Lattice.
MMLab , CUHK
6
0
9
4
Individual token selection and metadata caching unlock sparse attention for diffusion transformers without sacrificing generation quality, achieving up to $2.32\times$ speedups.
Visual perception agents can achieve significant performance gains by actively seeking information beyond a single glance, outperforming static models by up to 30 points on challenging benchmarks.
High-fidelity 3D texture generation is now achievable without any 3D assets, by cleverly transforming 2D images into effective training data.
Allocating additional computation during decision-making boosts long-horizon robot manipulation success by significantly enhancing subtask prediction accuracy.
GCache achieves a remarkable 2.17x speedup in video diffusion while enhancing visual quality, challenging the effectiveness of traditional caching heuristics.
