Skip to Main Content
Table 2: 

CiC-Zero Shot Generalization. Zero-shot generalization to unseen objects on the Chairs-in-Context (CiC) dataset. Results suggest NES can learn words as event classifiers in a general, object-agnostic manner. *SG model from (Achlioptas et al., 2019).

ModelZero-Shot ClassesAll
LampBedTableSofa
Major. 0.333 0.333 0.333 0.333 0.333 
*SG 0.501 0.564 0.637 0.536 0.560 
PoE 0.422 0.466 0.587 0.483 0.490 
NMN 0.462 0.492 0.572 0.532 0.515 
MAC 0.533 0.531 0.632 0.551 0.567 
NES 
w/VGG16 0.544 0.578 0.693 0.588 0.601 
w/Res101 0.573 0.589 0.715 0.610 0.622 
ModelZero-Shot ClassesAll
LampBedTableSofa
Major. 0.333 0.333 0.333 0.333 0.333 
*SG 0.501 0.564 0.637 0.536 0.560 
PoE 0.422 0.466 0.587 0.483 0.490 
NMN 0.462 0.492 0.572 0.532 0.515 
MAC 0.533 0.531 0.632 0.551 0.567 
NES 
w/VGG16 0.544 0.578 0.693 0.588 0.601 
w/Res101 0.573 0.589 0.715 0.610 0.622 
Close Modal

or Create an Account

Close Modal
Close Modal