EURASIP Journal on Audio, Speech, and Music Processing

Table 3 The results of YuYin that eliminating modalities on Commercial-98K

From: YuYin: a multi-task learning model of multi-modal e-commerce background music recommendation

Methods	Video \(\rightarrow\) music				Music \(\rightarrow\) video
Methods	R@10	R@15	R@20	R@25	R@10	R@15	R@20	R@25
YuYin pure	0.370	0.400	0.424	0.443	0.345	0.378	0.400	0.417
YuYin w/o E	0.374	0.403	0.430	0.453	0.345	0.377	0.404	0.424
YuYin w/o T	0.368	0.406	0.445	0.477	0.358	0.390	0.405	0.428
YuYin	0.403	0.439	0.471	0.501	0.376	0.423	0.456	0.478

Back to article page