Search papers, labs, and topics across Lattice.
1
0
2
RefCaptioner not only outperforms existing models in video captioning but also enables precise grounding of visual elements to multiple reference images, enhancing factual accuracy.
