Fig. 9From: Accurate visual localization with semantic masking and attentionPerformance results of challenging images. a Some representative query images that contain interfering objects. b Stacked bar charts of position error and orientation errorBack to article page