Fig. 5From: Accurate visual localization with semantic masking and attentionElement-wise multiplication of \({\mathcal{{I}}_i}\) and \({\mathcal{{M}}_i}\)Back to article page