Skip to Main Content
Table 7: 

Multilingual image-text retrieval results (in mR) on Multi30k. Models are trained on the English data. Avg results of non-English languages.

ModelendefrcsAvg
CEM3P (Ni et al., 2021) 86.0 48.8 39.4 38.8 42.3 
 
BEM3P 81.3 52.4 49.7 39.6 47.2 
CEM3P 84.2 52.6 49.6 33.4 45.2 
Sep+CoopM3P 84.4 55.6 52.2 39.8 49.2 
Joint+CoopM3P 83.5 54.2 48.4 39.4 47.3 
ModelendefrcsAvg
CEM3P (Ni et al., 2021) 86.0 48.8 39.4 38.8 42.3 
 
BEM3P 81.3 52.4 49.7 39.6 47.2 
CEM3P 84.2 52.6 49.6 33.4 45.2 
Sep+CoopM3P 84.4 55.6 52.2 39.8 49.2 
Joint+CoopM3P 83.5 54.2 48.4 39.4 47.3 
Close Modal

or Create an Account

Close Modal
Close Modal