Search papers, labs, and topics across Lattice.
Dawar Khan, Omar Mena, Donggang Jia, Alexandre Kouyoumdjian, Ivan Viola are with King Abdullah University of Science and Technology (KAUST), Saudi Arabia. E-mail: {xinyu.liu, dawar.khan, omar.mena, donggang.jia, alexandre.kouyoumdjian, ivan.viola}@kaust.edu.sa. Xinyu Liu is with King Abdullah University of Science and Technology (KAUST), Saudi Arabia, and also with University of Electronic Science and Technology of China, Chengdu, China. D. Khan and X. Liu are joint first authors with equal contributions.Manuscript received MM dd, YYYY; revised MM dd, YYYY. (Corresponding author: Dawar Khan.)
2
0
5
0
Control both multi-subject identity and multi-granularity motion in video generation with DreamVideo-Omni, a framework that uses latent identity reinforcement learning to avoid identity degradation.
Forget cloud costs: this benchmark reveals the real-world performance tradeoffs of running LLMs directly on XR devices like Apple Vision Pro and Meta Quest 3.
