Search papers, labs, and topics across Lattice.
Gaoling School of Artificial Intelligence , Renmin University of China , Beijing , China
3
0
6
5
LLM performance on new tasks can be boosted by intelligently reusing past computations, even when tasks are independent.
FATE not only retains temporal information but also encodes synchronization in a reusable embedding space, outperforming existing models in both semantic and temporal tasks.
By jointly training a keyframe sampler with an MLLM, MSJoE achieves state-of-the-art accuracy in long-form video understanding while significantly reducing computational cost.
