Error Understanding On Cub 200 2011 1

Average highest confidence (EfficientNetV2-M)

Average highest confidence (MobileNetV2)

Average highest confidence (ResNet-101)

Insertion AUC score (EfficientNetV2-M)

Insertion AUC score (MobileNetV2)

Insertion AUC score (ResNet-101)

Résultats

Résultats de performance de divers modèles sur ce benchmark

Nom du modèle	Average highest confidence (EfficientNetV2-M)	Average highest confidence (MobileNetV2)	Average highest confidence (ResNet-101)	Insertion AUC score (EfficientNetV2-M)	Insertion AUC score (MobileNetV2)	Insertion AUC score (ResNet-101)	Paper Title
Grad-CAM++	0.2659	0.3462	0.2647	0.1605	0.1284	0.1094	Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networks
SMDL-Attribution (ICLR version)	0.3306	0.5367	0.4513	0.1748	0.1922	0.1772	Less is More: Fewer Interpretable Region via Submodular Subset Selection
Score-CAM	0.2403	0.3141	0.2510	0.1572	0.1195	0.1073	Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks
HSIC-Attribution	0.2679	0.2914	0.2493	0.1611	0.1635	0.1446	Making Sense of Dependence: Efficient Black-box Explanations Using Dependence Measure

0 of 4 row(s) selected.