In this work, we enrich landscape and genre paintings by spatializing sounds for the drawn objects and scenes, which expands visitors' perception of the paintings and immerses them in the depicted scenarios. Plus, we personalize such spatial audio perception based on visitors' viewing behavior by applying gaze tracking. Through a preliminary user study with 14 participants, we observed that the gaze tracking-based audio augmentation helped people better focus on the areas of interest in the paintings, and enhanced their overall viewing experience.