Weakly Supervised Object Localization on ImageNet-1K (test)

68.7MaxBoxAccV2 (Mean)

Feature Direction Alignment + Consistency with Attentive Dropout

Updated 4d ago

Evaluation Results

Method
Feature Direction Alignment + Consistency with Attentive Dropout 2022.04	68.7	86.7	71.1	48.3
Feature Direction Alignment + Consistency with Attentive Dropout 2022.04	66.6	84.8	69.2	45.9
Ki et al. 2022.04	65.2	84.3	67.6	43.6
ADL + IVR 2022.04	65.1	-	-	-
HaS + PaS 2022.04	64.6	-	-	-
CAM 2022.04	63.7	83.7	65.7	41.6
ADL 2022.04	63.7	83.6	65.6	41.8
ADL + IVR 2022.04	63.7	-	-	-
HaS 2022.04	63.4	83.7	65.2	41.3
CALM 2022.04	63.4	-	-	-
SPG 2022.04	63.3	83.9	65.4	40.6
CutMix 2022.04	63.3	83.7	65.2	41
CALM 2022.04	62.8	-	-	-
HaS + PaS 2022.04	62.1	-	-	-
Ki et al. 2022.04	61.3	81.5	63.2	39.4
HaS 2022.04	60.6	80.7	62.1	38.9
CAM 2022.04	60	81	62	37.1
SPG 2022.04	59.9	81.4	62	36.3
ADL 2022.04	59.9	80.8	60.9	37.8
CutMix 2022.04	59.5	80.3	61	37.1