Abstract: Despite the success of multimodal learning in crossmodal retrieval task, the remarkable progress relies on the correct correspondence among multimedia data. However, collecting such ideal ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results