Fig. 6From: Accurate visual localization with semantic masking and attentionIllustration of pose regressionBack to article page