Search papers, labs, and topics across Lattice.
KAIST AI
3
0
6
Acceleration techniques for diffusion LLMs can unlock significant inference speedups, but only if we understand the intricate trade-offs at play.
Iterative refinement using complementary acoustic and semantic clues boosts sound scene segmentation performance beyond baseline metrics.
Slapping a simple, training-free module onto your object detector can dramatically cut down on false positives in fixed-camera scenarios, without sacrificing speed.
