Search papers, labs, and topics across Lattice.
Technion -Israel Institute of Technology
3
0
5
Text-to-image models may only need basic word meanings and order, not complex contextual embeddings, to produce high-quality images.
Forget finetuning: DynaEdit unlocks complex video edits like action modification and object insertion, all without training, using clever manipulation of pretrained text-to-video models.
Forget monolithic LoRAs: LoRWeB dynamically mixes a basis set of LoRAs to unlock SOTA generalization in visual analogy tasks.
