1 paper · 1 filter
Zhitao Liu, Zengyu Liu, Jiwei Wei +4
3D cross-modal retrieval is gaining attention in the multimedia community. Central to this topic is learning a joint embedding space to represent data from different modalities, su…