Experimenting with image based location reasoning using architectural cues

I am building an experimental AI tool that analyzes images to suggest real world location by detecting architectural and design elements and explaining why those cues point to a specific place.

I tested it on a public image with a known location and recorded a short video showing the reasoning process. The output was close but imperfect, which is expected at this stage.

I am mainly interested in whether explanation driven reasoning makes these systems more useful and interpretable.